If this info is valuable to you or someone you might know, feel free to share it.

Share

Good Morning,

Editor’s note: This month full disclosure I’m having a family health issue, so I have less time for research and writing than my usual. I’m running a limited time discount in case you want to support my work and financially anchor my ability to continue doing this, you can upgrade here. Due to these personal circumstances, the publication may include more guest contributor posts in October than usual. Grateful to have the support.

Subscribe now

Nathan Benaich and Air Street Capital released the 9th edition of their incredible State of AI 2026 report. I’ve been covering this every year since I started AI Supremacy because it’s a great snapshot and reminder of where we are at with some unique takes.

Explore the Report

Many of the charts today will come from this source.

Claude Haiku 5.5

Five weeks from the Anthropic IPO, Anthropic released their latest iteration of Haiu 5.5. Claude Haiku 5.5 is of course their cheapest, fastest, and most capable small model where they claim it costs around 75% less to run than Claude Haiku 4.5. Haiku 5.5 is considered to be the world’s frontier model in its class.

Only one Frontier Model in Reality

Nathan in the State of AI report insinuates that Google and OpenAI share the “frontier lab” placement with Anthropic, but I do not believe this is the case. Anthropic going IPO for me is the sole frontier model builder now in the world weeks to months ahead of all other AI labs.

I can understand The State of AI’s thesis of three labs sharing the frontier though:

The State of AI report is backwards looking and is written weeks before the present.

In Pioneering Computer Use Haiku 5.5 shows the Frontier is just Anthropic

Full disclosure Nathan Benaich is a British VC, so their report will have an AI accelerationist bias and may not report on social, public and civic issues that are important for the macro context. However for academic and industry synthesis, it’s a highly respected report:

Download the PDF

Air Street Press have a Newsletter, have made decent picks in the development of Generative AI and AI startups in general. Nathan also has a 33 minutes video review of the State of AI 2026 report you can check out on his most recent article on Substack. I have no affiliation with them, some of their best bets include:

Venture Capitalists Claim AI has moved beyond Benchmarks

• Benchmarks built to last years they say, are now saturating in months. It’s not entirely clear how to evaluate this.

Drama between OpenAI and Mathematics Community Continues

“We urge mathematicians to discontinue their work with OpenAI and to return to a vision of science that centers human understanding.”

The backlash to AI is only just beginning.

Terry Tao via the Association for Human Mathematics Communications Working Group. Read the full statement.

Read the blog here. This is a response to OpenAI exaggerating its capabilities when on October 6th, 2026, OpenAI published a GitHub repository (openai/math) containing 722 mathematical manuscripts (grouped into 372 subject families) tackling open research problems and long-standing mathematical conjectures.

AI’s role in the future of science is open for debate. OpenAI stated that out of roughly 4,000 posed problems, each generated paper consumed an average of about 3 hours of "ChatGPT Pro thinking compute".

OpenAI has been Caught Over-Stating its Revenue

A report by the Financial Times indicated that OpenAI’s annualized revenue as of late September 2026 stood at approximately $50 billion, which was roughly $20 billion lower than the $70 billion figure that had previously circulated in financial media. So we know that A significant portion of OpenAI's enterprise revenue comes through cloud distribution partners like Microsoft, Amazon, and Google.

According to FT reporting, the discrepancy arose from attempts by OpenAI's own investors to produce a direct comparison with Anthropic's annualized revenue. Annualized revenue run rate (ARR) ​is sometimes misleading ​sales metric, ⁠which often involves multiplying one month's revenue by 12. It’s not clear what led OpenAI to exaggerate its financial progress to profitability.

Company Behind Hermes

Nous Research may become a serious player in the personal agent B2C space for individual developers and AI first adopters. They have raised $90 Million in a Series B, but also have an Enterprise Hermes on the way. This is an incredible sign for the capability and future of open-source agents. The round was led by Robot Ventures, with participation from other backers including Nvidia, Union Square Ventures, Menlo Ventures, Samsung, and 1789 Capital, where Donald Trump Jr. is a partner.

Nous Research’s will begin a more organized push into the enterprise sector with “Hermes for Businesses,” allowing companies to deploy customized AI agents that can handle multi-step workflows while keeping their data private and secure. Nous was formed in 2023. There’s been a sudden influx of funding for personal assistants since the popularity of Meta’s Muse. Including Underdog (via Conway Research), Hark and Tab.

Muse has 2 Million DAUs

These are estimates from Similarweb, when Meta only had 3 millions weekly users just a few days earlier according to The Information based off of internal Meta data. It’s hard to tell how well Muse app downloads are translating into MAUs, let along weekly or daily active users. The Muse product appears in 2026 far less sticky than Chatbots were in 2023.

State of AI 2026 Report

Hard to believe that this is the 9th edition of State of AI by Nathan Benaich. View the full report.

Will Agents be more popular than Chatbots?

OpenAI’s study of agent adoption finds the fastest growth among non-developers, including people in legal, sales, recruiting, and marketing. Obviously OpenAI has been pushing Codex as its 1.2 billion weekly users of ChatGPT have more options. Outside of individual plans, there are around 9 million corporate/business seats across ChatGPT Team, Enterprise, and Edu offerings. The adoption of Codex however is key to its ability for sustained Enterprise revenue.

Is AI accelerating AI?

AI is already helping build better AI. According to Anthropic’s internal index, Claude led 26% of measured model R&D work in August, up from under 1% in February. How fast will internal model RSI teams accelerate new more powerful models?

Researchers set the tasks and supervise execution, and Anthropic reports that this is speeding up development.

Did Physical AI just Reach its GPT-2 Moment?

Skild’s S1 performed unfamiliar tasks from video demonstrations without updating its weights. With 100,000 hours of pre-training, it scored 66% on the company’s cumulative per-step success measure, against 9% for a language-prompted baseline trained on the same data and compute.

I have written about this trend already here.

If 2026 is around GPT-2 level, when do we arrive at GPT-3 or GPT-4 for robotics and humanoids with general purpose utility.

AI-assisted drug discovery is finally reaching late-stage clinical development

AI-assisted drug discovery is finally reaching late-stage clinical development. Generate:Biomedicines’ GB-0895 for severe asthma and Insilico’s rentosertib for idiopathic pulmonary fibrosis are recruiting in Phase 3. Enveda’s ENV-294 is in placebo-controlled Phase 2a trials for eczema and asthma. Human trials are of course the ultimate test and will determine whether these medicines are safe and effective.

Do this give AI in drug development a more viable path? Will Anthropic continue going further into biology, computational and synthetic biology and biotechnology? It appears they are interested in doing so. While projects like Biohub are expanding, even as China moves ahead in some aspects of biotech and new drugs.

The Inference Explosion

As the demand for compute accelerates but the cost per token is declining rapidly, inference has become the significant intersection of revenue and commercial capture. Meanwhile the cost of Capex and datacenters has risen tremendously due to HBM costs and other supply bottlenecks. This creates a severe environmental press and bottleneck for how fast and meaningful compute has become while AI sentiment continues to degrade both from ordinary American consumers and in politics. As models become more capable with declining prices, Chinese open-weight models have gained market. This is probably the biggest single that has occured in 2026 in AI.

Can Anthropic and/or OpenAI continue their 3x Revenue Growth?

With Anthropic going IPO in five days, they have shown incredible ARR Growth. We will soon have more details about via their S-1 that will go public a few weeks before the date. As for the first period of 2026, OpenAI and Anthropic’s combined reported annualized revenue run rates reached an astonishing $105B by late summer, up from about $30B at the start of the year. That’s more than 3x.

Research firm New Constructs wrote in a note on Tuesday that Anthropic’s IPO would be “the most ridiculous IPO of 2026,” posing a historic risk to financial markets. With a $2 Trillion valuation, we will soon find out what the market thinks.

I don’t believe personally that Google or OpenAI can be characterized as frontier model builders any longer and that Chinese open-weight models should take more token marketshare in 2027. "The Big Short" investor Michael Burry highlighted the valuation gap between Anthropic and established S&P 500 companies, noting that Anthropic's valuation equals the combined value of 78 S&P 500 companies, including top household brands and established industrial firms. A lot is riding on the IPOs of Anthropic and OpenAI (presuming OpenAI does go public in the first half of 2027).

Anthropic’s premium high-spend token share should continue as the leading frontier and Enterprise AI company in 2027, but I don’t believe OpenAI can keep up. OpenAI had to hit pause on Astra 6.1 due to alignment and cybersecurity issues. That is totally unprecedented for a company losing as much money as OpenAI continues to do. This is not what super intelligence looks like! I never trust any post lumping these two companies together.

The SaaSApocalyps Scare - real or imagined?

The success of Claude Cowork and its business plugins across legal, finance, marketing and more brought whiplashing uncertainty to software markets. Aided by some fairly deceptive macro (Citrini was later sold to Semianalysis along with its x following) coverage. Nearly $285B was wiped from software stocks in two weeks during February’s “SaaSpocalypse.” This thesis that AI will be disruptive to SaaS companies continues to persist, when while at the same time I will note that SaaS companies are some of the best at executing new high-growth products related to LLMs.

The Problem of Local Consent for the AI Buildout

With a significant rise in State wide datacenter moratoriums, AI is becoming a political issue. More AI requires financing, construction, and local consent. But these are not infinite where in 2026 AI sentiment sourced considerably for most people. This as a European plan is seeking up to €10B in public funding and at least €20B privately. While IPOs like Nscale will test “Sovereign compute” in places like Norway. Most funding of European AI, robotics and datacenter initiatives actually comes from the U.S.

RESEARCH

Read more

It’s impossible to cover AI and not talk about Meta Muse. This is not an endorsement, just some of my observations and information. Just yesterday Meta has announced Meta Enterprise Platform.

This at a time when AI has literally disrupted media and creatives1.

AI and the Fall of the Creative Class

Read It

Young people might be deskilled in how to code and how to write simultaneously, within just a micro generation. That’s not by the way, the empowering AI we were promised.

Will Agentic Swarms Disrupt Core Skills of Human Literacy and Intelligence?

ChatGPT Era is official Over

Whether we like it or not, agentic swarms are here. In terms of AI Supremacy on the consumer (B2C) front, the last four years have been the Chatbot era where ChatGPT of OpenAI was the dominant leader2 (late 2022 to late 2026). This weekend + all Monday I’ve been going over Meta’s Muse product and I can say with some confidence that the ChatGPT era is now really over. We’re about to embark for likely the next few years into a journey of Personal Super-Intelligence Agents (PSIA). Meta is SOTA in PSIA, let me explain:

Subscribe now

Muse’s Rapid Ascent and a new Personal AI Race

I predict Muse will have around 1 million users by November, 2026. In 2027 it could continue to grow quite fast. The AI industry is a bit confused about the branding, not a “companion”, but maybe more like a teammate or casual personal assistant that is supposed to help you optimize your personal life that will become increasingly capable to actually accomplishing tasks for you.

When do Personal Super-Intelligent Agents Arrive?

The new era of AI begins when two things occur in Meta’s timeline: the arrival of its Watermelon model (due in late October) and the finishing of its Prometheus (near Columbus in New Albany, Ohio) datacenter coming fully online in the next few months. By then adoption of Meta’s Muse agent will be even more impressive than its first month. So this new era arrives circa or around December, 2026. Right after Anthropic’s IPO a new race in consumer AI kicks off (that’s right after U.S. midterms). Simultaneously ChatGPT will be losing marketshare to open-source models that are taking more token marketshare especially in the last six months3. OpenAI is thus in deep trouble in 2027 for a few major reasons:

Signs of Cracks in OpenAI Dominance are starting to show

1. Pressure in Consumer AI from Meta and SpaceXAI.

2. Pressure from Chinese open-source models on cost and token marketshare.

3. Late to IPO, and markedly over-taken in ARR and growth by Anthropic in Enterprise AI.

With Google late to deliver their next frontier model, among BigTech incumbents only Meta has made a serious challenge to be a frontier player in both LLMs and consumer AI products. The Meta AI app even before Muse has gotten very decent traction the past nearly 1.5 years. Muse is currently available only in the U.S. and Canada.

Try Muse

Meta’s Scale Advantages Up-sell its First Comer Personal Super-Intelligence Move in Muse

Muse was launched on September 8, 2026 or around 3 weeks ago. By the time its Watermelon 🍉 model drops it will be about seven weeks and where more countries will likely have been rolled out to try the first iteration of Muse. Meta has two major advantages over SpaceXAI, OpenAI and others trying to be players in the next personal AI market. These are huge numbers of daily active users and a warchest of Ad revenue. Meta reports an average of 3.60 billion “family daily active people” (DAP). Meta is also the internet’s dominant digital advertizer.

Meta’s full-year total revenue is projected to be in the area of $245 billion and Muse could materially accelerate their advertising spend for small businesses at scale. Meta also has resources to spend on marketing their new AI products that most AI startups don’t have. Meta is widely expected to have significant negative FCF by the end of 2026. The compute spend on Muse will make a difficult company to value, and likely far more risky to invest in. All of this inspite of Meta’s impressive adoption rates and the performance of the app on leaderboards.

Token Spend and Machine-to-Machine Inference (M2M Internet)

The PSIA (personal super-intelligent agents) market will radically accelerate token use and only truly profitable companies will be able to compete due to the high compute (costs to subsidize), talent, operations, training and RSI costs. These systems will be designed to keep getting better and more personalized to the user including up-selling them to new more specialized AI products in the next few years and nudging them to use Meta’s digital advertising to grow their side hustles, entrepreneurial projects or existing small businesses.

Products like Muse will contribute to a new definition of what it means to be compute constrained in the late 2020s. It’s going to be so expensive, few will be able to follow.

Share

Why is Meta Muse the First Mover?

Meta’s Muse is the consumer culmination of Meta’s ambitions in the Metaverse and AI merged. According to my calculation Meta will have spent upwards of $180 Billion on this PSIA project by 2027. They will have gone into significant debt to do it.

That’s multiples more than Grok Bot of SpaceXAI or OpenAI’s Codex Bot (“o”) have to put into a persistent agent experience with dedicated memory and a cybersecurity hardened and encrypted Virtual Machine space for swarms of agents to work dedicated for you. Meta also has such a treasure trove of data to enable Muse to help us maintain and “show up” for our friends, family and personal connections. This is language Mark Zuckerberg specifically used in his multiple recent interviews and podcasts.

It’s actually fairly interesting to think of Muse as the first major project of Meta’s new Meta Superintelligence Labs founded around 15 months in.

Meta’s Muse Timeline

It only took AI to go from ChatGPT to Muse in 45 months. Not fast at all when you think about it. ChatGPT was released on November 30, 2022, and Meta's Muse was released on September 8, 2026. But for consumer B2C AI products, a lot happend in those nearly 1,400 days. For most people it hasn’t been that exciting to say the least. Meta’s Muse might have a great team, and the product might still not catch on like ChatGPT did four years ago. But how did we get here?

Meta’s Superintelligence Lab (MSL) was formed just around 14 months ago: Meta’s core Superintelligence frontier team has about 44 all-star members. They are supported by thousands of other machine learning researchers. Muse is the result of a long journey of capital, compute and talent concentration (that will be hard for Google, OpenAI or anyone else to emulate). The entire MSL division is around 3,000 AI researchers with a total of around 6,500 AI related employees in the applied AI support unit. This is what it takes to build something like Muse.

• Some Context: The early prototype of OpenClaw was first launched on November 24th, 2025. OpenAI later acqui-hired the founder on February 14th, 2026 but failed to capitalize on the opportunity. The entire time Muse was in development.

• April, 2025: The major setback for Meta's Llama series occurred with the release and aftermath of Llama 4 in April 2025, culminating in early 2026. This signalled Meta AI would need a restart.

• April 29th, 2025. Meta releases its standalone Meta AI App on April 29, 2025

• Late June, 2025. The Great talent poaching: The bulk of Meta's aggressive poaching to build its Superintelligence Labs took place during the summer of 2025, specifically kicking off in June and July.

• Late June, 2025. Mark Zuckerberg announced the creation of Meta Superintelligence Labs (MSL) on June 30, 2025, with the official public announcements and restructuring details following in early July 2025.

• Late July, 2025. Meta's blog post titled "Personal Superintelligence" (also published under the title "Personal Superintelligence for Everyone") was written and published on July 30, 2025, read it here.

• Mark Zuckerberg's supposed Public Manifesto, a 6,000-word essay titled “The Future is for Everyone” was released on August 10, 2026. Read it here.

• Meta’s Muse agent was launched on September 8, 2026

• Meta’s next-generation frontier AI model, codenamed "Watermelon," is expected to be released in late October 2026 and to be at the level of GPT-5.5. approx.

• Re Prometheus datacenter near Columbus Ohio, the marquee milestone for Prometheus to hit its initial 1-gigawatt (GW) target is scheduled for late 2026 around December, 2026. Major data halls and initial on-site gas-turbine power infrastructure (such as the Socrates sites) are turning online throughout late 2026 to reach this capacity. (Hyperion is approx 5x the size of Prometheus).

• Meta’s biggest datacenter Hyperion, will likely finish its first initial stage in 2027. The first data halls and infrastructure zones are scheduled to complete commissioning and go online, delivering an interim capacity target of roughly 1.5 GW in Richland Parish, Louisiana. Full capacity will take at least until 2030.

• September 28th, 2026 - Meta has just announced Meta Enterprise Platform.

• September 29th, 2026 - Meta announce Muse for Small Businesses.

Is Muse a ChatGPT Killer?

I’m curious if these persistent AI agents will essentially cannibalize vanilla chatbots. How about for you in the time you spend with and on AI?

What can you really do with Muse? (Use Cases)

Read more

Welcome Back!

Visions of AI is a new feature format I’m experimenting with that will amount to a short profile on an AI or emerging tech startup. The cadence of this style of article is unknown as of yet, but there are a lot of fascinating startups I want to discuss and share about.

This is designed to be light evening reading to go out at a time-slot of 8 pm EST.

In an AI boom you would expect new kinds of AI startups to emerge and get funding. While China has what it calls the world robot conference, to complement the vast limitations of large language models (LLMs), another kind of model is being pushed forward - the elusive world model.

Subscribe now

A Starting definition of a World-Model

“A world model is an internal mental simulation constructed by an AI system that represents the physics, spatial relationships, geometry, and dynamics of the real world.”

As we build the physical infrastructure to power the compute for some much debated and uncertain AI-future, all these robots and IoT systems needs a better understanding of their environments.

Given the limitations of LLMs, it’s not clear how long or how useful these physics based and simulation friendly world models might take to be able to fast-track the training of our suddenly humanoid looking robotics projects.

While China has established itself a formidable open-weight model builder, that western companies are increasingly building upon (like Harvey building a model on a Moonshot AI base), the new race to world models is heating up.

As a Canadian it was exciting to hear that Khosla Ventures and Radical Ventures are leading the seed funding of what it turns out to be one of the largest seed financings ever raised by a Canadian startup. It’s called Veeda and is by three former Nvidia employees with ties to the University of Toronto.

Veeda Innovation will focus on building multi-modal world models.

View Mission

Embodied intelligence and physical AI are really hitting a feverish pitch in the late summer and early Fall of 2026. On August 19th we learned that renowned computer scientist Sanja Fidler, is setting up shop with former Nvidia colleagues and raising $90 million in a seed funding round today. Sanja is a Slovenian-Canadian AI researcher held in very high regard.

Read Introduction

There are now around a half dozen very good world model startups with extremely good funding.

Share

“We started Veeda because we believe robotics will reshape the world—changing how we move people and goods, how we manufacture and build, and how we operate in the physical world.”

The publication called The Logic was one of the first to break the story. If you are a robotics enthusiast this is very exciting. As Anthropic sets for a record breaking IPO there’s some indication it has serious interest in the field and perhaps even acquiring Physical Intelligence.

Helping Robots Learn and Train Faster

Read more

Good Morning,

Visions of AI is a new feature format I’m experimenting with that will amount to a short profile on an AI or emerging tech startup. The cadence of this style of article is unknown as of yet, but there are a lot of fascinating startups I want to share about.

This is designed to be light evening reading to go out at a time-slot of 8 pm EST.

Not that many AI researchers have worked at Google DeepMind, OpenAI, xAI and started their own companies. Igor Babuschkin in 2026 is the founder of River AI. River is building a stack for personal AI that is both open-source and adapts to you. This is a bit like the holy grail of personalization combined with all the recent advances in AI.

I actually find myself attracted to the philosophy and goal of the startup:

We imagine a future where your AI works entirely for you and deeply aligns with your values. Compared to the corporate chatbots of today, it will feel completely different: responsible, curious, and truly yours. Most importantly, you'll own everything: the hardware it runs on, the data it learns from, and the intelligence itself.

This has legit open-source vibes to me and the interview with the lead investor General Catalyst felt authentic:

River AI is Hiring

The interview is with Hemant Taneja, the CEO of General Catalyst, one of the lead investors of the Seed/Series A.

Chapters
02:22 — Co-Founding xAI
04:36 — An AI Experience You Own
05:59 — Betting Early on Open Weights
07:47 — How Do You Make Money on Free Models?
09:25 — Why the US Needs Its Own Open Models
11:27 — Enterprises Want to Control Their AI
13:36 — What Is Continual Learning?
16:56 — An API That Updates the Weights
19:17 — A Personal AI for Everyone
20:46 — The 20-Minute Training Run

A 2-Month old Startup

River AI is among a group of AI labs that are taking on hard problems with quick capital not seen previously in the history of Venture Capital. It’s also very clearly a bet on Open-Source American AI.

Subscribe now

Read more

👋 Hey there, I’m Mike. Each week I share AI articles at the intersection of tech, business, society and the future. If you want to support the channel or gain full-access to my work, go here. Read Archives | See Substack Notes | Visit our community Chat | Visit Homepage. A 24 year old Hedge Fund star has emerged and become more famous by the boldness of his mistakes. The AI theme may feature briefly in this story.

Get Full Access

Good Morning,

I just wanted to share a little note. Or if you prefer to listen. (20 minutes, 19 seconds).

How should we understand the cautionary tale of AI investor and Hedge fund manager Leopold Aschenbrenner in his rapid rise to fame and fortune? For context AI investor Leopold Aschenbrenner worked at the FTX Future Fund (the philanthropic arm of Sam Bankman-Fried's cryptocurrency exchange FTX) before it collapsed. He later joined OpenAI's Superalignment team, and after leaving, founded the AI hedge fund Situational Awareness LP. He had a knack for spotting “bottlenecks” in the Semiconductor industry and predicting them well before stocks surged.

Who intersected Sam Bankman-Fried and Sam Altman in quite the same way while heralding the advent of AGI? I remember reading his Situational Awareness manifesto and wondering who was funding him? Remember Sam Bankman-Fried and Alameda Research led a Series B/early-stage investment into Anthropic in 2021 when the AI startup was valued at roughly $2.5 billion. There are a lot of Anthropic connections in this story. That eight percent Alameda stake would be worth $77.2 billion today.

As you’ve likely already heard, Leopold Aschenbrenner’s hedge fund was forced to sell all of its public stock holdings where Ken Griffin’s Citadel hedge fund swooped in and reached a deal to buy the stocks. Of course now he wants to raise even more capital. Situational Awareness, which grew to as big as $45 billion at one point, posted large losses in recent weeks from declines in AI infrastructure investments like SK Hynix but he still has a huge amount of Anthropic shares and made some decent bets, if a bit concentrated.

He was born in Germany to parents who were both doctors. At age 15 he convinced his parents to let him move alone to the United States and enroll at Columbia University. At Columbia he double-majored in economics and mathematics-statistics. He co-founded the university’s effective altruism (EA) chapter, conducted research connected to the Global Priorities Institute at Oxford, and co-authored work with economist Philip Trammell. At 17 he received a grant from Tyler Cowen’s Emergent Ventures; Cowen later described him as an “economics prodigy” after reading one of his papers. He graduated as valedictorian in 2021 at age 19.

In 2023 he joined OpenAI’s Superalignment team (led by Jan Leike and Ilya Sutskever), which focused on technical approaches to controlling AI systems more capable than humans. He co-authored the paper “Weak to Strong Generalization.” In April 2024 OpenAI fired him; the company cited an alleged information leak, while Aschenbrenner has disputed the characterization and pointed to internal security concerns he raised (including about potential foreign espionage risks). The Superalignment team was dissolved shortly afterward.

In a very short time, now 24, he’s become one of the most tracked Financial Creators on X. Aschenbrenner launched the fund after leaving OpenAI in 2024 and quickly became one of the most watched figures in AI investing because of eye-popping returns, according to CNBC.

In June 2024, roughly two months after leaving OpenAI, he self-published the ~165-page essay Situational Awareness: The Decade Ahead. It argued that AGI could arrive as early as around 2027, sketched a path toward superintelligence, highlighted compute/energy constraints, geopolitical risks (especially involving China), and national-security implications. The piece circulated widely in AI, tech, policy, and investor circles and various AI boosters.

Read Manifesto

His Fund has done fantastically well, where returns since inception had exceeded 1000% and the fund had expanded to around $20 billion, according to the FT. Rumor has it he has to sell about half of his Anthropic shares in the recent Bear market correction on Semis of late July, 2026.

The hedge fund Situational Awareness LP (named after the essay), initially with seed capital of roughly $225 million from backers including Nat Friedman, Daniel Gross, and Stripe co-founders Patrick and John Collison. These are very powerful and influential Angel investors in the AI space. Notably Jane Street, the Quantitative trading firm appears to be a major backer. Jane Street has also been incredibly successful in this AI bubble/boom. It’s a very mysterious connection for me.

Leo became a bit of a cult figure around investors cashing in on the Semiconductor boom. But his success is also going to lead to more success, even though his calls have been incredibly high-risk:

The fund hit $45 billion at the start of July. His biggest bets were SK Hynix, Nebius, SanDisk, Micron, and CoreWeave, all down more than 35% this month (at the time of writing). His shorts on software names like Adobe went against him too. Citadel bought the bulk of what was left. He's 24 or maybe just 25 years old.

Read more

Good Morning,

With NASA’s Artemis II mission (April 1st to 11th, 2026) going around the Moon right now, there’s a lot of buzz for the aspirational Lunar economy that’s coming. But what does thi…

Read more

Good Morning,

I’m really bullish on the future of the Voice AI interface.

• Once agentic AI and AI wearables mature, that dream of “ambient computing” is going to be closer to reality for many consumers.

• Generative AI is enabling new convenience for consumers that will reshape how we access services, find information and deal with recurring tasks.

• This has big implications in both B2C and B2B.

AI Wearables will transform Voice AI Convenience

I believe once better AI wearable devices are launched next year in 2027 including AI pins (wearables) and smart glasses form factors, it could be a breakthrough year for consumer AI voice experiences. From smart glasses to new kinds of pendants and pins, I expect Apple to dominate. Mid to late 2027 is the time this should really come to the foreground or around 18 months from now. For details about Apple’s upcoming AI devices read here.

To wrap your head around the ecosystem:

Voice AI Experiences will improve dramatically in 2027

Smart glasses are going to be good enough to strike a chord with consumers with strong offerings of both AI (and later AR) capabilities from Meta, Google, Alibaba, Xiaomi, Xreal, RayNeo (雷鸟创新) and of course Apple, among others.

The interface of how we interact with AI will become more multi-sensory, hands-off and accessible. The dream of ambient computing has a lot of potential to evolve in the decade ahead.

Since the below infographic was made, there’s been dozens of AI Voice agent startups rewriting the future that are scaling now. We’ll list some of the major ones later on. They begin to mature in the late 2020s.

I asked of Why Try AI Newsletter to help me us understand Genspark’s new voice agent in Genspark Workspace 2.0. They have even introduced AI calling partnering with Twilio. This is the point in the week where I think we need to summarize some AI news events too of the last couple of weeks.

✨ Your AI News Brief 💼

Around 9 min.

Key Trend Analysis among Generative AI Apps and AI Tools

Recently ’s Olivia Moore released their Top 100 Gen AI Consumer Apps, a series that I’ve been following religiously. While Claude saw a spike recently, it still only has about one 20th the daily active users of ChatGPT.

Olivia Moore Breaks down the AI Consumer Highlights

March 11th, 2026.

Voice AI News

• Claude Code rolls out Voice mode capabilities.

• ElevenLabs released Scribe v2.

• Read Claude Voice Mode.

• Promising Voice AI startup Hume AI was gobbled up by Google.

• Voice AI agents are going vertical.

ChatGPT’s Marketshare is Eroding

While OpenAI’s ChatGPT still has a huge base, Gemini, Grok, Claude and others are gaining marketshare too. A trend that appears to be accelerating.

Why Try AI

Hand-on AI for non-techies. Stay grounded with the latest AI tools. By

Why Try AI

Hype-free, hands-on AI for non-techies.

By Daniel Nest

• Here Are My Go-To AI Tools

• Sunday’s AI News Rundown

• Build Your Second Brain With Claude Code & Obsidian

Consumer Paid Subscriptions 2026

I expect Gemini, Claude and Grok to have big years in 2026 and 2027 regarding consumer premium subscriptions putting pressure on OpenAI and slowing down their ARR (revenue growth). This is Anthropic and Cursor (Anysphere)’s Enterprise ARR keeps accelerating.

I’ve been impressed by the pace of innovation at Genspark and while I miss their Perplexity like AI search product, they are also entering the voice AI picture.

Genspark Workspace 2.0: What’s The Big Idea?

of Why Try AI takes a deep dive into Genspark’s evolution of their AI Workspace and their Voice AI capabilities in particular. With explainer videos.

Since we wrote this piece Genspark has already released Workspace 3.0 just yesterday. Mere months after its Series B funding and “unicorn” status, Genspark launched the latest iteration of its agentic concept: AI Workspace 2.0.

AI Workspace 2.0 brings improvements to several existing agents as well as new features:

New launches:

• Speakly: A new dictation assistant (think Wispr Flow and Superwhisper) that you can download for macOS or Windows. It lets you interact with all of Genspark’s agents and tools via voice instead of typing.

• AI Music Agent that can create custom music (think Suno).

• AI Audio Agent that does voiceovers and narration (think ElevenLabs).

Upgrades:

• AI Inbox now supports automated workflows that perform specific actions like creating daily inbox digests, interacting with external messaging platforms like Slack, analyzing social media performance, etc

• AI Creative Slides, AI Image Agent, and AI Video Agent are all more capable and incorporate the improved powers of newer, better underlying models.

While the under-the-hood upgrades and new music/audio agents are neat, what Genspark is leaning heavily into is Speakly and the promise of hands-free agentic work. Simply say what you need done, and Genspark’s Super Agent does it.

At least, that’s the idea.

To better understand Genspark’s vision, I tested the three new features and also conducted a written interview with the company’s COO, Wen Sang.

Read on to find out if Genspark can be more than the sum of its parts.

Testing the new tools

Let’s start with my hands-on tests and demos of the three new features.

1. Speakly: The voice agent

On the surface, Speakly is yet another voice dictation app. You speak into the microphone, and text comes out on the other end.

But what makes Speakly a great fit for Genspark’s infrastructure is the “agent” shortcut that sends your spoken request directly to Genspark’s Super Agent and the baked-in, intelligent processing of spoken input.

I demonstrate Speakly’s main features in this 7–minute hands-on video:

Watch it to learn:

• How regular voice dictation works

• How Speakly can auto-correct filler words and backtracking

• Agent mode: Send a request or task to Genspark Super Agent from any screen

• Translation mode: Speak in any language (or combination of languages), get English text.

• Custom modes: From “Buzzwords” mode to “Twitter” mode, Genspark can process and rework your speech into any style or format. You can customize this to your needs.

If you want to try Speakly for yourself, grab it here:

[DOWNLOAD SPEAKLY FOR MAC OS OR WINDOWS]

2. AI Music: From requests to music tracks

If you have toyed with AI music sites like Suno or Udio, you already have a good idea about how this works. The agent can use different third-party AI music models to execute your task:

But because Genspark agents can coordinate their work with other agents, you can have more elaborate requests that require additional pre-processing. So I tried this:

Prompt: Watch this video about the launch of Genspark AI Workspace 2.0 and create a grand soundtrack worthy of this announcement:

Ed: All to say Genspark has a lot of exciting new products and features (not a Sponsor). Just a really innovative case study that I like to cover since nobody else is.

Try Genspark

Voice AI is Evolving in the 2020s

Where are we seeing concentrations of voice agent startups and products? 🤔

• Customer Support

• Sales & Lead generation

• Recruiting & Human Resources

• Finance

• Insurance and Legal

• Logistics

• Home Services & Small Business Operations

• Healthcare

• Personal Voice Agents

• In this article the Genspark deep dive continues with more Voice AI insights and infographics.

Subscribe now

Read more

Thank you for your support on my last article: The U.S. Labor Market and its AI problem. I’ve been writing about AI for well over four years, well before the latest Generative AI craze. A special thanks to Michael Burry who shared the piece. It made my week.

Good Morning,

While I am a contrarian and realist in my coverage of AI as my baseline, there are however times when my own exuberance and optimism for a better path in AI research shines through. An alternative to the deceptive AGI marketing (by Sam Altman) is one of those moments.

To be sure Generative AI has ushered in an era where Artificial intelligence has made substantial progress over the past decade (2017 to 2022 being a fairly nascent period). Predictive and generative systems have transformed how we analyze, retrieve knowledge, and create content at global scale. But this isn’t enough. The Machine Economy ahead will demand more.

Like many of you I’ve been following the criticism around LLMs by the likes of Yann LeCun, Gary Marcus and many others in the last few years. Yann LeCun, a Turing Award winner and a pioneer of modern AI, has become one of the most prominent critics of the current "LLM-centric" path and his alternative is to me fascinating. Instead of AGI they propose Superhuman Adaptable Intelligence, or SAI.

“SAI points toward self-supervised learning for acquiring generic knowledge from unlabeled data, and world models for planning and zero-shot transfer.” - Ravid Shwartz Ziv

Read the Paper

The reality is we need a second wave of more practical AI startups that focuses on Physical AI, world models, robotics and real automation and that’s exactly what we are seeing with Prometheus Project, Core Automation, World Labs and many others. AGI startups, World model centric AI startups, Physical AI & robotics startups and physical science AI startups all can push the frontiers.

In this article I explore what the Machine Intelligence Economy will require and in particular try to chart the origins, characteristics and announcement of Europe’s own AMI. I argue that we are moving into a new era of Physical AI, specialized AI startups, robotic brain and world model origin labs. I consider 2027 as the nascent start to the Physical AI era.

“AI agents need world models that allow them to predict the consequences of their actions before they take them. This is key to enabling agents that can plan, remember, and reason about complex observations.” - Michael Rabbat

There’s now in 2026 a second wave of AI startups that are offering complementary directions to the approach of LLMs.

Subscribe now

Europe’s Biggest AI Startup Came out of Stealth

Read more

Just a reminder I’ll be working on my Emerging Tech Investments Newsletter a lot more in 2026 to discover speculative high-growth and buy the dip opportunities. The cadence of that Newsletter should increase in the weeks and months ahead. There I’m exploring themes like Neo Clouds, Space-tech, micro cap AI related names, Quantum computing SPACs and more.

Good Morning,

This is going to be a long report with a lot of infographics and context.

Analyzing Friday jobs data, nonfarm payrolls fell by 92,000 in February, according to the Bureau of Labor Statistics, far below consensus expectations of roughly 50,000. It was the third time in five months that the economy lost jobs in the U.S. Outside of Healthcare jobs, there’s really not much growth at all with low hiring. We are in a “jobless” growth economy, with K-shaped characteristics.

With the U.S. putting the breaks on immigration and Tech companies doing agentic AI pilots, more layoffs in the Tech sector are highly likely. Oracle needs to shed at least 30,000 jobs due to the debt they have taken on for OpenAI’s compute facilities. Even as OpenAI’s Stargate facility won’t be expanding any time soon.

A Low Hire Economy with Grave Consumer sentiment

• Hiring is down 20% lower compared to the pre-pandemic baseline of 2019 according to Karin Kimbrough of LinkedIn, their head economist.

• People are staying unemployed for 7 months on average.

Labor market impacts of AI

AI is Lowering Entry Hiring in the Most Exposed Professions

Anthropic economists say they have developed a new way to track how AI will upend the workforce.

Their economists Maxim Massenkoff and Peter McCrory are selling a pre-IPO pitch, that exposed knowledge workers could be more impacted by the technology that Anthropic building. As has pointed out, they are pulling old data and making some huge rationalizations. The likes of Anthropic and OpenAI say the red will inevitably fill the blue. But how likely is that really?

According to Anthropic, AI is far from reaching its theoretical capability: actual coverage remains a fraction of what’s feasible. Occupations with higher observed exposure are projected by the BLS (Bureau of Labor Statistics) to grow less through 2034.

While the circle graph went viral on social media, the bar graph is an easier way to visualize this:

Outside of coding, and some aspects of administration and finance, Generative AI doesn’t seem very capable of automating tasks in most areas of knowledge work. The deeper problem appears to be that the tech is not a meaningful job creator or increasing productivity outside of a few roles even in technology companies. While massive layoffs are not occuring, some redesigning of jobs internally at tech companies have been noted. Less managers, more hybrid roles, more product managers and designers doing vibe-working on the job.

Hiring slowed in January, 2026 significantly with U.S. hiring falling 3.3% from December and is down 5.7% compared to January 2025. - source. Hiring in 2025 was already anemic.

Generative AI does not appear to be a Job Creator

The GDP gains of the investments going to datacenter, inference and compute campuses and building isn’t being spread equally in the U.S. economy. The low-hire labor environment has other issues as well.

Monthly U.S. Job Creation is Nearly Non-Existent outside of Healthcare in 2025 & 2026

• Healthcare is the main driver of new jobs created in 2026.

• Cognitive displacement, labor talent and youth deskilling (College) and “cognitive surrender” are now AI risks are that are moderate to high in such an environment.

A broader measure of unemployment that includes discouraged workers and those holding part-time positions for economic reasons moved lower, to 7.9% or 0.2 percentage point below the January level. Where the 4.4% unemployment rate is deceptive, compared to the low-hire environment job seekers are facing.

Then there is the AI pressure on hiring. Young people have reason to be more nihilistic or discouraged with regards to the current U.S. labor market trends.

Generative AI is Destroying White Collar Opportunity

Read more

Good Morning

Recently with the Anthropic push back against the U.S. DoD (“Department of War”) issue, a group of quantum scientists have independently published a manifesto rejecting the use of quantum research for military purposes and is seeking signatures from researchers around the world. Read the manifesto. You want to talk about emerging technology? The worlds of National defense, Generative AI, Geopolitics, Robotics Innovation (including things like space-tech & drone swarms) and Quantum computing are quickly converging.

Quantum computing is still in its nascent emerging beginnings. But it’s slowly becoming a field worth following with potential major intersections with AI, hybrid quantum chips, and very specialized accelerators for very narrow tasks related to scientific frontiers like chemistry, new materials, battery tech, cybersecurity and national defense capabilities, among others.

The Three Core Pillars of Quantum

The three rapidly (Q3) emerging pillars of Quantum technology and quantum computing broadly are as I see them are quantum computing, quantum communication, and quantum sensing that could together generate up to $97 billion in revenue worldwide by 2035. This could be around $200 Bn. by 2040.

1. Quantum Computing

2. Quantum Communication

3. Quantum Sensing

I’ve always considered Quantum computing a wild card in how AI Supremacy plays out. I’ve been covering Quantum computing startups, fundraising and the industry for years on my Quantum Foundry Newsletter. When we finally are able to have real machines with millions of qubits (far from today’s reality), it will vastly opens up new possibilities for computing. It’s going to be a journey to get there.

Quantum computing will lend aspects of parallelism that could radically augment AI. Even in terms of Quantum machine learning at the intersection of LLMs, if you think about it, thanks to superposition, a quantum computer can evaluate millions of model parameters simultaneously rather than sequentially. Companies like Nvidia have released NVQLink (a so-called "Rosetta Stone" of hybrid computing), which allows GPUs and quantum processors to communicate with microsecond latency. This hybrid hardware will evolve considerably in the next decade.

A Period of R&D could lead to a Quantum Acceleration period

Classical AI struggles with high-dimensional data and our future AI systems will have access to new emerging computing architectures and paradigms. QML or Quantum machine learning is a new frontier where hard engineering and hybrid utility are evolving steadily in the 2020s. Just as we have frontier labs and AI startups to use AI to design our own chips, so will we will one day I predict also have this for Quantum computers, new kinds of hybrid chips and entirely new computing paradigms. This is not well understood by those in AI, Venture Capitalists or even National tech policy think tanks today. The way Quantum and AI evolves together will enable an altered future that is highly uncertain. Quantum computing even in its infancy is today impacting finance, cybersecurity, advanced materials, battery tech, biotechnology, drug development, logistics & supply chains, among many others.

Qubit Modalities and Quantum Hybrid-Chips

Qubit types are still evolving rapidly although a real breakthrough in Quantum remains elusive. Some say that Neutral atoms (using laser-cooled atoms like Rubidium or Cesium) are arguably the most promising for massive scaling in the next 2-3 years. Meanwhile watch the R&D Nvidia is doing with TSMC. TSMC are owning the infrastructure that are attempting to create a Quantum Bridge. TSMC R&D means TSMC is producing the specialized Nvidia Quantum-X and Spectrum-X photonics switches using their leading-edge nodes (3nm and 2nm) for advanced quantum nodes.

More Quantum Companies going to the Public Market

With yet another Quantum startup going public this February in the SPAC of Infleqtion (ticker ), investor awareness about Quantum computing will only continue to grow. Quantum, like AI (Anthropic vs. Department of War), is thus becoming a National Defense technology of considerable importance. In cybersecurity, communications and sensing with ramifications in space-technology as well. Also in predicting certain outcomes in high-powered simulations.

I asked for his take on where the Quantum computing industry is at, so our AI readers can catch up and be current. Brian writes:

Quantum's Business

Quantum Technology for Business from a Global Quantum Top Voice and LinkedIn 'Top Strategy Voice'.

By Brian Lenahan

Just as National Defense spending increase for space-tech, some of that might also boost Quantum startups. This is because things like cybersecurity and Quantum sensing have mission-critical implications. The most serious Quantum startup I’m waiting for to go public is Quantinuum. Meanwhile on January 14, 2026, its majority owner, Honeywell, announced that Quantinuum has filed a confidential draft registration statement (Form S-1) with the SEC for an Initial Public Offering (IPO). In Quantum, Europe is not seen as a laggard like it is in scaling major Tech companies and AI, but as a pivotal player. It makes the whole space rather interesting to watch.

Quantum Companies and Qubit Modalites to Watch 2026-2030

The Venture Capital scene behind Quantum is fairly interesting, as are the areas where China is competitive or even ahead. There’s certainly been some moonshots in Quantum startups, but the future is still fairly uncertain. Government funding of the Quantum industry for National Defense is also a huge driver as well as the impact from huge corporate sponsors.

Most Well Funded Quantum Companies 2026

Some of the leaders in terms of funds raised in recent years are:

• PsiQuantum

• SandboxAQ

• Quantinuum

• IQM Quantum Computers

• Xanadu

• QuEra Computing

• Multiverse Computing

• Classiq Technologies

• Alice & Bob

• Pasqual

Xanadu going public

On January 28th, 2026 Xanadu based in Canada also took steps to begin the process of going public. When Quantinuum and Xanadu go public, they together with IonQ will represent the most promising first three Quantum companies on public markets. This will also mark a period where BigTech will allocate more capex in Quantum and making more strategic acquisitions.

Quantum Modalities

A brief overview:

Topological Qubits

Microsoft has also been working hard on a Modality that is a Moonshot where in brief they are trying to build Topological Qubits using Majorana quasiparticles.

Best Funded Quantum Startups

Not all Quantum modalities of qubit approaches have a valid or scalable future. Right now the Neutral Atom and Trapped Ion approaches seem dominant. But that could change.

The State of Quantum: Brian Lenahan

Brian is the founder and chair of the Quantum Strategy Institute (QSI), a collaboration of quantum experts and enthusiasts from around the globe enabling business to understand the technology, its potential and its practical applications.

He’s a Quantum author, strategic analyst and expert I’ve been following for years. He’s a consultant, mentor, think tank leader, workshop facilitator, as well as a 3x Amazon Bestselling author, public speaker and business leader in the space. He has a Quantum strategy for business course here. One of his most reviewed books was called Quantum Boost (2021). If you are curious about the history of Quantum computing read this one (2023).

Where are we today in 2026 in Quantum?

By of Quantum Business Newsletter.

In late 2024, I sat down with the quantum leadership team at Microsoft’s Seattle campus along with a group of industry analysts and participated in a compelling focus group followed by a lab tour. The previous year, I had walked through the D-Wave quantum lab in Burnaby, BC to see a quantum computer or “fridge” up close. In Boston, I toured the QuEra lab which includes a lego version of their quantum computer. In fact, I’ve watched the evolution of quantum computing with a mix of cautious optimism and relentless scrutiny as a former bank executive who focuses squarely on results.

So, I have been somewhat amazed and thrilled about 2025. 2025 stands out as the year the field decisively shifted from “promising lab demos” to “credible paths toward practical utility.” Designated by the United Nations as the International Year of Quantum Science and Technology, the recognition amplified global attention, inspired billions of dollars in investment, and research breakthroughs.

What is Quantum Computing?

If you’re an AI enthusiast, think of quantum computing as the next leap beyond classical GPUs and TPUs (Tensor Processing Units). Today’s AI models crunch massive data with billions of parameters using classical bits (0 or 1). Quantum computers (QC) use units in the form of atoms or particles called qubits (or quantum bits) which can be 0, 1, or both at once (thanks to the crazy world of physics called superposition)—like exploring many possibilities simultaneously (think running through a maze in every direction at once rather than a linear, one path-at-a-time approach, to solve the maze faster). QC’s also exploit entanglement, where qubits link such that changing one instantly affects another, no matter the distance. This enables exponential speedups for certain problems that classical computers (even the biggest supercomputers) struggle with, like simulating large molecules for new drugs or optimizing complex AI training for financial portfolios or large cities.

Classical computers will never be replaced by quantum computers because they’re used for different purposes. Think of it this way – today’s computers are good at analysing large amounts of data with few parameters in such functions as accounting, operations, basic drug design and testing, whereas quantum computers are best with smaller datasets but a large number of parameters (or complexity). So, just the hardest problems would leverage a quantum computer.

What’s New?

The Evolution of Sensing

As an industry writer, observer and conference presenter, I have the good fortune of connecting with many of the leaders of quantum companies, and their persistence (and now access to greater funding through both private and public capital) has translated into significant advances. And before you mention the oft-heard predictions of quantum technologies being years or decades away, I point you to one pillar - quantum sensing - in navigation especially – which is already in market today, replacing jammable GPS tech, as an example of quantum’s commercial progress.

If you’ve been on a commercial flight where GPS signals were jammed—leaving the aircraft reliant on less precise backup systems—you’re likely feeling uneasy about aviation safety in an era of increasing electronic interference. But imagine your plane is equipped with Q-CTRL’s Ironstone Opal quantum navigation system. Your confidence would quickly return, because this advanced quantum sensor technology provides positioning that is completely passive, undetectable, and inherently immune to jamming or spoofing. Unlike traditional GPS, which depends on vulnerable satellite signals, Ironstone Opal uses ultrasensitive quantum sensors—enhanced by proprietary AI-powered software—to map subtle variations in Earth’s magnetic field (or gravity in related implementations). This geophysical approach delivers GPS-like accuracy without any external signals that adversaries can disrupt. Real-world flight trials have demonstrated it outperforming high-end conventional inertial navigation systems by up to 94x (or more in some cases), ensuring reliable, secure navigation even in fully GPS-denied environments. In short: when GPS fails, quantum-assured navigation doesn’t just keep you on course—it restores peace of mind.

Quantum and AI

Quantum technologies and AI do indeed have direct relationships. Quantum speedup could dramatically reduce time and energy consumption for training massive LLMs or foundation models (e.g., hours instead of weeks), while many AI tasks (hyperparameter tuning, combinatorial problems in logistics/scheduling, portfolio optimization) map to NP-hard problems where quantum algorithms like QAOA (Quantum Approximate Optimization Algorithm) or quantum annealing show promise. In reverse, AI and Machine Learning is proving essential for building and operating quantum systems, which are noisy and hard to control including error correction and mitigation where neural networks decode error syndromes and reinforcement learning discovers better codes. The combination also can optimize pulse sequences to minimize noise (e.g., Google’s and others’ work on AI-driven quantum control) and AI can aid in discovering new materials or architectures for qubits. Finally, AI can help design better quantum circuits/algorithms.

The Evolution of Computing

The ultimate quantum computer is one with the accuracy of today’s systems (99.99999999999% or better) yet being able to handle much more complex problems (such as mentioned optimizing traffic in a large cities or a financial trading portfolio) with substantially more variables or parameters. It’s true the best today’s QC’s can manage is 99.99%, yet that level is vastly improved from just 24 months ago. And while collectively we are not yet at large-scale, everyday-useful quantum computers, 2025 delivered verifiable milestones in some of the priority challenges fixing errors, scaling hardware, and showing real advantages over classical systems.

The Big Fix: Error Correction Becomes Real (The Key Breakthrough)

Quantum bits are fragile—tiny disturbances like heat, vibrations, or cosmic rays cause errors quickly (referred to as ‘decoherence’). This has been the biggest industry roadblock for years. In 2025, the field crossed a major threshold with quantum error correction (QEC) moving from theory to hardware reality. Google’s Willow chip (105 qubits, superconducting type) achieved the “below-threshold” milestone. By grouping many physical qubits into one reliable “logical” qubit and using clever codes (like surface codes), errors dropped exponentially as more qubits were added. Willow ran a benchmark task in about 5 minutes that would take the world’s fastest classical supercomputer 10^25 years—way longer than the universe’s age. More excitingly, they demonstrated Quantum Echoes, the first verifiable quantum advantage on a real, useful algorithm (out-of-order time correlator), running approximately 13,000 times faster than classical methods. This wasn’t just “faster for fun”—it ties to problems in physics, finance, and potentially AI pattern recognition. IBM advanced with processors like Quantum Loon (testing fault-tolerant parts) and Nighthawk (high-connectivity for complex circuits). Their roadmap targets Quantum Starling by 2029 where 200 logical qubits are expected to be running 100 million error-corrected operations. Microsoft, the same organization I visited, pushed topological qubits (Majorana 1 chip) for built-in error resistance, and novel 4D codes reduced errors dramatically in simulations. To boot, research exploded with 120+ peer-reviewed QEC papers in the first 10 months of 2025 (up from 36 in 2024). Error correction shifted from “maybe someday” to “now an engineering challenge,” meaning bigger systems get more reliable, not less.

Hardware Progress: Many Approaches Racing Forward

Akin to Beta versus VHS in the early era of videotape, or Android vs iOS in today’s world, many “modalities” exist in the world today though no single “best” way to build qubits has emerged. Modalities like superconducting, trapped ions, neutral atoms, photonic and annealing types of quantum computers exist simultaneously. Superconducting computers from companies like Google, IBM, and Rigetti operate like tiny loops cooled to near absolute zero (colder than outer space). Willow and IBM’s Heron showed high fidelity (99.99%+ accurate operations) though nowhere near classical fidelity of up to 15 nine’s. Trapped ions & neutral atoms computers from vendors like Quantinuum, IonQ, and Atom Computing function with ions/atoms held by lasers for high accuracy. Quantinuum’s Helios (launched late 2025) claimed the most accurate commercial system, enabling generative quantum AI. IonQ hit advantages in drug discovery and chemistry simulations. Photonic computers from companies like PsiQuantum use light particles to perform computations. Annealing computers predominantly from D-Wave (my other visit mentioned above) specialize in optimization problems.

Early Wins: Quantum Advantage in Real Applications

2025 saw credible “quantum advantage”—quantum outperforming classical on narrow but useful tasks such as IonQ/Ansys achieving 12% better medical device simulations, Google achieving 13,000x speedup on verifiable algorithms, and Quantinuum & others achieving better accuracy in chemistry, materials, and AI-related tasks. These target drug discovery (simulating molecules exactly), climate modeling, finance optimization, and even enhancing AI (e.g., better randomness for secure models or quantum-inspired training). Quantum sensing (as mentioned above) also progressed starting with MRI’s decades ago to ultra-precise, non-jammable magnetometers for navigation.

Investment, Ecosystem, and Reality Check

Funding surged in 2025 with $3.77B in the first nine months (nearly 3x that of 2024) including PsiQuantum’s $1billion and $800 million to Honeywell-owned Quantinuum. Governments poured in billions with Japan leading the way in 2025. Cloud platforms (IBM, AWS Ocelot, Azure) made experimentation easy. Certain challenges do remain including significant talent shortages (not just PhD’s), high energy/cooling needs, and full fault-tolerance for broad use still 5–10+ years away. But 2025 proved scaling is feasible and physics isn’t the blocker anymore.

Scale & Democratizing Quantum Access

Read more

AI Supremacy is a Newsletter about AI at the intersection of business, innovation, technology, society, and the future of civilization. Follow me as Read Futurist.

Seventy-five years ago, the idea of harnessing the power of the skies was little more than fantasy spun by futurists like Arthur C. Clarke and Isaac Asimov.

What if, we are about to see it happen in our generation.

~

Good Morning,

This piece is going to read a bit different from usual pieces, because it’s a topic I’ve been considering for some time (where I cover space-tech companies and Neo Cloud observations).

This weekend (basically the last four days) I’ve been frantically pondering how scaling AI requires not just a better semiconductor supply-chain, but a radically improved energy source and cost-efficiency structure. We have the nascent technology to manifest this already. But as a civilization are we able to do it? Who are the major players going to be?

On the topic of orbital datacenters it’s worth considering a more futuristic solution to AI Infrastructure and energy-efficiency and that to me might be trans-orbital manufacturing, Space-mining and scaling AI with energy of the Sun. It’s very science-fiction like pre 2026, but not as far off as you might think.

This is because rocket launches get cheaper and far more numerous and the energy bottleneck becomes far more intense for AI and capex in the decade ahead due to new (fairly dirty and inefficient) datacenter projects. My base case is that lunar and orbital bases will become accelerated, space manufacturing becomes normative, and new concepts around AI datacenters in space will proliferate. Even in the late 2020s.

It won’t be enough to spend a $Trillion dollar1 on capex (like they will in 2027), it will require a radically more energy and cost efficient solution to scale.

The Catalyst

When SpaceX goes public in an IPO later in 2026 (maybe as early as June), the promise of the future isn’t just about datacenters, it’s about harnessing more of the Sun’s energy to fuel AI at a sale that isn’t currently possible with the U.S. energy grid constraints, or even terrestrial datacenters.

Share

Easily one of the most important Elon Musk interviews ever, Read the transcript here, we are going to be going over this quite a bit.

Watch it on YouTube

The Rise of BigAI

Imagine a future where OpenAI, Anthropic and SpaceX (where xAI is now merged as a subsidiary) all become key AI Cloud computing full-stack companies. The one with the most compute will have an absurd advantage over the others, if you believe scaling AI with more compute will matter.

Let me repeat, they will all become Cloud computing hyperscalers in their own right too. Orbital datacenters and future space infrastructure affords an AI Cloud on a planetary level of scale. 🌍 There is no decent concept of this in 2026, it is a pioneering landscape.

AI will press new Cloud Computing companies into existence:

Read more

Good Morning,

As some of you know I’ve been following the Semiconductor industry pretty seriously for the past few years. While TSMC has pulled ahead of the competition in recent years from Intel and even Samsung, there is one future contender I’m watching that I’m bullish on and that’s Japan’s Rapidus.

I’m of the belief that to understand AI you have to have a good understanding of the semiconductor supply chain, key players, and an appreciation of how important not just datanceters are, but the hardware that enables GPUs. Their niches and their stocks. Rapidus is easily among my favorite global semi startups. Many people who follow AI have never even heard of them. They plan to go IPO that is targeted for fiscal year 2031.

Rapidus Corporation is a Japanese company dedicated to the research, development, design, manufacture and sales of advanced logic semiconductors, in their own words. Essentially this is one of the most exciting Semiconductor chip startups (not a sponsor). Rapidus is racing towards 2nm mass production and it’s super important for the future of AI globally. I don’t find Rapidus is getting enough credit.

The Top TSMC Challenger in Asia

Rapidus Corporation is a Japanese semiconductor company founded in August 2022 to develop and manufacture advanced logic chips (2-nanometer node) by 2027. Backed by major Japanese firms (Toyota, Sony, NTT) and government support, it aims to establish a high-performance foundry for AI, autonomous driving, and edge computing to boost Japan’s competitiveness in the chip industry.

The Semiconductor industry is undergoing a generational boom with even Elon Musk confirming that Tesla plans to build a "TeraFab" in the U.S. to manufacture its own AI, logic, and memory chips, reducing reliance on suppliers like TSMC and Samsung. With Huawei racing to catch up to Nvidia, this is all shaping up to be a dynamic decade for the future of the Semiconductor industry that is also experiencing an HBM chip shortage.

There’s a Significant Semiconductor Boom

Prices for computer memory, or RAM, are expected to rise more than 50% this quarter compared to the last quarter of 2025. Three primary memory vendors that are Micron, SK Hynix and Samsung Electronics — make up nearly the entire RAM market, and their businesses are benefiting from the surge in demand. Micron’s stock is up 39% just this year and a stunning 387% over the past year.

While Elon Musk’s so-called future TerraFab will combine memory, chips and packaging in one facility claiming it’s the only way to reach the needed scale, I see the rise of Rapidus in the next five years as a distinct possibility. I do believe TSMC is a monopoly and for good reason, where Rapidus is one of the best newcomers to the space in years. How they utilize AI in their business model is also super relevant.

I asked (Starr) a talented entrepreneur, student and writer out of Philly to dig into it for us in more detail. Jason is the founder of one of the fastest rising Semiconductor Newsletters on Substack and says he’s a chronically online college student obsessed with semiconductors posting deep single-name equity research because it’s fun.

Jason takes you on deep semiconductor tech journeys via 50+ page research reports in as entertaining and simplified a way as possible.

I highly recommend you follow his Newsletter.

Jason's Chips

I take you on a journey from deep semiconductor tech to DCF-derived intrinsic value via 50+ page research reports in an entertaining and simplified way.

Why Rapidus Fascinates me?

Rapidus could be the most intriguing cinderella story in the entire semiconductor industry. Learning about the semiconductor industry is also all about learning about business in places like Taiwan (where I live), South Korea and Japan. That’s why I started the Semiconductor Things Newsletter.

• Rapidus is Japan's state-backed "all-or-nothing" bet to reclaim its crown in the semiconductor industry.

• Rapidus encompasses an enormous amount of collaboration among Japan’s top companies and Government.

• Founded in 2022 by eight Japanese giants (including Sony, Toyota, and SoftBank), it aims to jump from Japan’s current decade-old technology directly to the world’s most advanced 2-nanometer (2nm) chips.

• So far so good: They broke ground on their IIM-1 fab in Hokkaido in 2023, completed the cleanroom in late 2024, and successfully prototyped 2nm Gate-All-Around (GAA) transistors in July 2025. Rapidus isn't inventing 2nm from scratch. They are using technology transferred from IBM (who built the world's first 2nm test chip).

Can Generative AI accelerate Semiconductor Fabs?

• In late 2025, Rapidus unveiled Raads (Rapidus AI-Assisted Design Solution). It uses Large Language Models (LLMs) to help engineers design chips specifically for the 2nm process.

Following the launch of Raads Generator and Raads Predictor, Rapidus will also release additional tools throughout 2026 including:

• Raads Navigator / Raads Indicator: Utilizes LLMs to provide quality assurance and assistance to designers to derive solutions to design issues and problems.

• Raads Manager: A layout design tool that utilizes ML/AI to create a hierarchical configuration that minimizes design time.

• Raads Optimizer: Applies ML/AI to search for and derive parameters to optimize PPA.

Jason’s Chips Publication

Jason’s work also is a must-read for investors covering a lot of the topics and companies I care about. Here’s why: his goal is to do technical engineering research to uncover alpha, build comprehensive operating models and DCFs, then distill his work into clear, cohesive, and hilarious stories readers intuitively understand. He models using a revenue build + 3 statements + DCF (Discounted Cash Flow) valuation approach with granularity beyond what you see on the sell side.

He’s doing a B.S. in Finance and Artificial Intelligence and I believe in emerging writers just like him. Unlike other tech/semiconductor investing Substacks he focuses a bit more on the financial side understandably. As a student he’s seeing all this with fresh eyes: He adds:

“Because after all, as much as we love the tech, we buy stocks for their free cash flows. So, you’ll hear a lot more about operating leverage, cyclicality, capital intensity, and accounting than you usually would from another semis writer.”

Pledge Jason Chips

He writes classic deep dives, portfolio reviews and even Earning reviews. There simply aren’t many writers around who can do that (3.9 GPA). Semiconductor and Investing Newsletters are among the most profitable on Substack and since he’s early his prices are still really low, so grab a paid sub to him while they are this low.

His body of work for just five months since he started his Newsletter is impressive.

Rapidus are targeting high-end, specialized AI accelerators, like a genuine national moonshot for Japan. Japan has come into focus in 2026 for multiple reasons, I don’t have to tell you we live in interesting times.

Japan in Context: 2026

Japan's benchmark Nikkei 225 stock average soared over 2,000 points to all-time highs. Japanese Prime Minister Sanae Takaichi's party is likely to score a landslide victory (above left).

The deep dive today on Rapidus is locked for premium readers. In February, 2026 free articles go out guaranteed on Thursdays. Title:

Rapidus: Japan’s Leading Edge Semiconductor Manufacturing Hopium

My deepest appreciation to Michael Spencer for giving me the opportunity to write this guest post. Please see my other content at Jason’s Chips. Author’s trending works:

1. The Lumentum Series | Part 1: Transceivers

2. The Lumentum Series | Part 2: Co-Packaged Omnipotence

3. The Aixtron Series | Part 1: Power Electronics

4. The Aixtron Series | Part 2: The Optoelectronics Supercycle

Japan’s National Security Imperative

Read more

This Newsletter is about AI Supremacy. Between the U.S. and China, but also among the biggest AI Startups. OpenAI in 2026 is likely to be pressured by both incumbents and AI startups, by Giants and more nimble specialist companies.

.

.

We’ve never seen anything like it: OpenAI can go bust and it won’t even matter… I can’t stop thinking about OpenAI’s path ahead and how it might bifurcate from the real winners of AI in the year ahead of 2026. Some field notes follow. No napkin math but some opinions that might be relevant.

• People don’t fully realize what an IPO frenzy 2026 is going to bring and how much capex, debt fueled AI infrastructure build-outs and private AI startups valuations will keep rising next year.

• Circular financing of the 2025 variety is only going to increase, because Silicon Valley is betting on Generative AI to increase profits and earnings. With so much liquidity on the sidelines (due to years of QE and pandemic mismanagement by the Fed), the AI boom has yet to fully manifest (on the stock market) as the bull market prepares for AI frontier model, Chip and Neo Cloud companies to go IPOs in 2026.

• IPOs, M&A and more AI startups with unrealistic valuations will mean a capital allocation binge like we’ve never seen before in the history of technology.

There is little signs of the AI boom slowing down - Hyperscale datacenter operators nearly tripled their spending on infrastructure over the past three years in response to the AI craze, while the amount of operational capacity added each quarter has increased by 170 percent.

OpenAI is in talks to raise up to $100 billion in a funding round that could value the ChatGPT maker at up to $830 billion, The Wall Street Journal reported last Thursday, however something doesn’t quite feel right. xAI, Thinking Machines and other AI startups founded by AI luminaries have equally unrealistic private market valuations.

With about one week to go in 2025, More than $61 billion has flowed into the data center market. So why would OpenAI be committing Trillions in future deals when its Enterprise, API and even B2C chatbot marketshare has gone down fast in 2025? OpenAI is also censoring its own research. Meanwhile Nvidia with over 90% of the AI Chip market and TSMC on the way to capture 50% of Foundry 2.0 revenue worldwide, OpenAI is not shaping up to be a monopoly leader in its own space of chatbots.

OpenAI’s momentum begins to enter real challenging competition in 2026 as more rivals emerge and more sophisticated ones evolve faster.

This article mostly discusses OpenAI’s prospects in 2026.

The OpenAI Downfall Setup

Read more

2026 will be a year of reckoning for Sam Altman, in more ways than one. He will be 41 in April, 2026.

Share

OpenAI is likely to do a historic Trillion dollar IPO in 2026 or more likely, in 2027. Six days ago OpenAI confirmed their conversion to a public benefit corporation PBC structure. A mega IPO was always part of the plan for the cash pumped and margins starved chatbot pioneer. OpenAI made a $12 Billion loss last quarter. Impressive. It’s Sam Altman, against the world. It’s all reminiscent of a Gary Marcus or Ed Ziton rant. But I’m not sure I consider these people serious analysts on the business side of things.

To date, OpenAI have raised an astounding $57.9 billion and the cash injection of an IPO is just what they need to stay above water. According to a Bain Capital Ventures partner1, for years, OpenAI has communicated like the CIA - mysterious briefings, leaks on X, and a general vibe of “trust us, it’s complicated.” Their PR was flashy and exaggerated, but lacked substance, details and a coherent plan.

Recently about five days ago, for an hour, Sam Altman and Jakub Pachocki (Chief Scientist) went live and laid out dates, dollars, and doctrine. It was equal parts revelation and fever dream and was an amusing projection forward:

So four days ago: (I’m not a huge OpenAI YouTube watcher but this one is worth it)

Sam, Jakub, and Wojciech on the future of OpenAI with audience Q&A

Watch the Video

In the video OpenAI explores a new “AI research intern” product that they hope will change the future. They claim that by Sep-26, they expect an “AI research intern” that meaningfully assists human scientists. By Mar-28, Sam Altman said it will be a “fully autonomous researcher.” Ambitious! And totally doubling down on their AGI marketing. But the infrastructure around OpenAI looks fairly dubious with some circular math. This AI Scientist plan reminds me of Tesla trying to spin the idea of Robotaxis as their salvation. That’s the problem with over-promising…

In this post I’m going to analyze some of the major issues and fun topics that jump out at me about OpenAI’s business future and its IPO. I’m both excited and afraid for them. Because when you crunch the numbers, it gets really entertaining if you are an AI nerd like me.

OpenAI is betting that consumer global adoption and AI Infrastructure can be a sufficient moat against competitors, but is that even going to be true? Scale itself might have been a moat in Peter Thiel early days but the AI economy has so much liquidity that no longer holds true.

OpenAI is going head-to-head with Apple, Google, Meta, Microsoft, xAI, Anthropic and China, all at the same time. Sam Altman is a decent venture capitalist, but he’s not a CEO that you’d want in a company prepping for an IPO. While Nvidia has rock-solid margins, OpenAI’s entire business model sounds somewhat speculative in a world where their ARR growth is likely to deaccelerate in 2026 or 2027 and margins burn on the fringes. Or will it? Depends on what else OpenAI as up its sleeve.

I run a frequent AI chat where I ask the community (all of my readers using the Substack app) for their real opinions, insights and quick takes:

Join the Conversation

Will OpenAI’s ARR Deaccelerate as Competitors Ramp up in 2026?

According to The Information, in Q3 2025 OpenAI projected its 2028 revenue to be $100 billion.

• The 2026-27 pivot is key and a successful IPO is crucial. OpenAI desperately needs an IPO and hopefully ASAP before they continue to lose more API and consumer market share.

Curious about my archives (going on 4 years now), getting all my work and the full-range of AI Supremacy coverage?

Get Full Access 🔑

Maybe we need to see the future differently? 🤔

Kindly share this note.

Now let’s dig into some of the details.

OpenAI as AI Infrastructure Giant?

Read more

How different are Thinking Machines Lab and SSI actually going to be from OpenAI and Anth…

Read more

I’ve been pondering the monstrosity that is Nvidia and Microsoft nearing the $4 Trillion market cap milestone as it marks a change to the future of capitalism and a radical leap in the future of business empires. The concentration of hyperscalers is now reaching epic proportions.

Nvidia is just run differently. I can’t stop thinking about what it means for the future of emerging technologies like general purpose robots and new architectures of computing like Quantum computers with millions of qubits.

In some ways, Nvidia isn’t just an accelerator of Sovereign AI for the various regions of the world, but an incubator for the emerging technologies of the decades ahead.

• Nvidia is the Fish that Learned to Drive a Bicycle

• Digit Gets Smarter: How Agility Robotics Is Leveling Up with NVIDIA

• 🏥 The Robotic Rounds Begin

• The Island of Healing Machines: Inside Taiwan’s Bold AI Healthcare Experiment

Nvidia’s $4 Trillion Milestone

Nvidia first crossed into the $4 Trillion market territory on July 9th, 2025.

Only 11 companies are in the $1 Trillion club, to put the achievement into perspective:

Most Valuable Companies by Market Cap

As of July 11th, 2025:

Gaming and Bitcoin Mining to AI and Synthetic Reasoning Inference

In a nutshell: Nvidia’s graphics processing units (GPUs), originally developed for gaming, became the backbone of artificial intelligence (AI) infrastructure due to their ability to handle the massive parallel computations required for AI model training and inference.

With hyperscalers building more datacenters to handle the demand for compute, the rest is history. But in the 2022 to 2025 period, a shocking history unfolded nonetheless.

Nvidia’s Revenue went from $4.4 Billion to $73.88 Billion in 2 Years

Nvidia’s financial performance has been remarkable, with net income soaring from $4.4 billion in fiscal year 2023 to $73.88 billion in fiscal year 2025.

More recently, first-quarter 2025 revenue reached $44.1 billion, a 69% increase year-over-year, driven by AI chip demand.

The company forecasted $45 billion in revenue for the second quarter, reinforcing investor confidence in its growth trajectory.

I asked to cover Nvidia’s incredible GTC event a few months ago. It was a pivotal moment for the company in their coming-out party.

Diana is my favorite blogger relating to Physical AI and robotics:

Introducing the 🤖 DROIDS Newsletter 🏜️

In 2025, the NVIDIA Jetson™ and Isaac™ platforms provide end-to-end solutions to develop and deploy AI-powered autonomous machines ushering in a new physical AI future.

• One day when we think about Nvidia, we might not just picture AI chips and datacenters, but the robots that help build the future of the world.

DL with the Wolf 🐺

DROIDS is the new Newsletter about robotics by .

DROIDS!

Riveting dispatches from the front lines of robotics.

By Diana from Droids

• Nvidia is the first company to ever reach a valuation of over $4 Trillion. That its an AI chip semiconductor company and not a software, Cloud or advertising company is telling.

• Nvidia CEO Jensen Huang met with the U.S. President Donald Trump at the White House on Thursday last week, a day before the chip designer's head plans to visit China. There are geopolitical layers and trade tensions and Nvidia and its products are at the center of it all.

• Nvidia’s mad-rush to the top has been legendary: The California-based company, which was founded in 1993, first passed the $2 trillion mark in February 2024, and surpassed $3 trillion in June. Both Nvidia and its closest peer AMD have Taiwanese-American CEOs (AMD: Lisa Su). Their most important partner is TSMC, the crown jewel of Taiwan’s towering semiconductor supply-chain.

• Broadly speaking, Nvidia has profited heavily from the growing demand for AI hardware and chips since the launch of ChatGPT in late 2022. Just three years later no one company in the world has as much leverage in the converging of emerging technologies with AI as does Nvidia.

While Nvidia’s GTC 2025 occured in March, it ushered in a more public demonstration that Nvidia is going to be important for the Physical AI, robotics and Quantum computing era. Nvidia’s software moat, partnerships and rapid revenue increases means it’s in a unique position to usher in the future of innovation beyond just AI chips and Autonomous Vehicles.

Let’s briefly look at some of the main revenue generating segments of the company:

How does Nvidia make Money?

Nvidia primarily makes its money through the design and sale of graphics processing units (GPUs), related hardware, and software solutions. Here’s a breakdown of its key revenue streams based on the latest available data:

1. GPU Sales for Gaming: Nvidia’s GeForce GPUs dominate the gaming market, powering high-performance graphics for PCs, laptops, and gaming consoles. This segment generates significant revenue from gamers and PC manufacturers.

2. * ✨ Data Center and AI: The fastest-growing segment, Nvidia’s data center business, includes GPUs (like the H100 and upcoming Blackwell series) optimized for artificial intelligence, machine learning, and high-performance computing. These are sold to cloud providers (e.g., AWS, Microsoft Azure), enterprises, and research institutions for AI model training and inference.

3. Professional Visualization: Nvidia’s Quadro GPUs and other professional graphics solutions cater to industries like architecture, engineering, media, and entertainment for tasks like 3D rendering and simulations.

4. Automotive and Embedded Systems: Nvidia develops chips like the DRIVE platform for autonomous vehicles and AI-powered automotive systems, sold to car manufacturers and suppliers. This also includes embedded systems for robotics and IoT.

5. Software and Services: Nvidia offers software platforms like CUDA, AI frameworks (e.g., NVIDIA AI Enterprise), and cloud-based services (e.g., DGX Cloud) for AI development and deployment. Licensing and subscriptions contribute to recurring revenue.

6. Networking: Through its acquisition of Mellanox, Nvidia sells high-speed networking solutions (e.g., InfiniBand, Ethernet switches) critical for data centers and AI infrastructure.

Watch the Keynote

• Datacenters and AI

• Physical AI and Humanoid robotics

• Quantum computing and World Models

was lucky enough to attend GTC 2025 on a press pass provided by NVIDIA, so this is coming from a first-person witness perspective.

GTC this year was a great reminder about what Nvidia is becoming and what it means for technology as a whole and the planet.

Skim the GTC 2025 Keynote

• What follows is Diana’s first-hand exploration of GTC, its implications and key impacts.

• With a special emphasis on Nvidia’s deep partnerships with humanoid robotics companies.

• Many analysts believe the ChatGPT moment for humanoid robotics and/or Quantum computing are in the area of 3-10 years from now, so roughly the 2028 to 2035 period. Nvidia’s role might be pivotal.

NVIDIA GTC 2025: The Dawn of Generalist Robotics

How NVIDIA's Latest Announcements Are Reshaping the Future of Physical Machines

The future greeted me just inside the door of the San Jose Convention Center.

Read more

Hey Everyone,

With uncertain macro conditions, AI startups and startups in general are shifting their strategies and building companies completely differently. But how? While I don’t write on Venture capital at the intersection of AI and startups often, it’s one of my favorite things to track as an emerging tech analyst.

The idea of seed-strapping and the dream of solopreneurs being able to scale startups in a more lean and agile manner with less employees with AI is fairly fascinating. New cases studies are emerging to inform the founders of today and the future.

In the era of Generative AI, the way founders and solopreneurs are bootstrapping is very different where there are many examples of AI founders who are able to scale revenue faster, be more agile and rely less on traditional equity dilution to grow fast in a more sustainable and in a less high-risk manner. Is this the beginning of a fundamentally different future of entrepreneurship with AI?

Let’s try and find out. You might have seen this quote:

“🚨 25% of today's YC Demo Day startups have codebases that are 95% AI-generated.”

Are smaller AI-augmented teams the future of agile startups? How would they function and raise funds differently?

What is Seed Strapping?

"Seed-strapping," or raising one round of funding and scaling profitably from there, is gaining popularity among startup founders amid a downturn in venture capital markets. Artificial intelligence is helping startup founders become more efficient for less money, thus minimizing the need to raise venture capital.

In the AI hype bubble, VCs are going after AI startups and throwing capital at them like never before. In early April, 2025 it’s come out that even a16z Andreessen Horowitz is reportedly seeking to raise $20 billion for a tech investment fund focused on growth-stage investments in American artificial intelligence startups.

Top Newsletters for Founders?

Now on Substack recently there are suddenly really great Newsletters at the intersection of Venture Capital, startups and AI for founders. Let me name some of them here (these are mostly ideal for founders, entrepreneurs and solopreneurs especially those who are looking for resources):

1. The VC Corner by

2. The Founders Corner by by

3. Product Market Fit by

4. a16z Speed Run by

• Product Newsletters that talk about AI are also highly recommended because they draw on tactics, strategy and product mechanics that enable startups and their products to grow.

• Tashi Dorjee from Sydney Australia has a really great YouTube introduction to what Seed Strapping is and why it’s becoming more popular here. Another great introduction to seed strapping is here.

However the founder that most impressed me with Seed Funding and his startup tips is actually . His viral posts on LinkedIn really caught my attention.

He’s a Waterloo alum who is technology entrepreneur, investor, and technologist, known for his work in AI, particularly for his role as an angel and seed investor in various AI startups, and for founding Super.com, which he grew to $150MM+ in annual revenue. Follow Henry Shi here on LinkedIn.

Henry’s Best Hits 🦄

If you are a founder, have an entrepreneurial itch or are a solopreneur you might want to follow his work.

Henry’s Best Hits

I help founders, write about AI and share behind-the-scenes lessons from founding Super.com ($150MM+ revenue/year). Raised $150MM in VC. Angel Investor.

By Henry Shi

• Seed-Strapping vs Boot-Scaling in the AI Native Era

• Launching the Official Lean AI Leaderboard 🏆

• Henry Shi’s AI Crash Course (Github)

• The Complete Guide to OKRs

The idea that AI is changing the future of startups, growth and entrepreneurship is really evolving fast especially as AI becomes better at coding and as AI agents begin to find frameworks like MCP by Anthropic and A2A by Google.

Breaking: Read the Blog

The best intro to MCP is by Latent Space via . If you understand what MCP and A2A evolve into the idea of AI empowered entrepreneurship is much bigger and more important trend than vibe coding.

As agentic AI frameworks and models get better, the function of entrepreneurship and scaling of revenue growth in startups also begins to changes.

Who’s Henry? 💡

Henry Shi is a successful founder and entrepreneur. He is currently writing about AI and helping founders build and scale businesses through angel investing. Paying it forward and giving back, one interaction at a time.

Let’s get into his incredible deep dive now:

4 Startup Funding Models in Age of AI

A guest post compilation by (March, 2025)

I recently exited my $150MM annual revenue startup that’s raised $200MM in venture funding and discovered something shocking:

The way 90% of founders build companies is fundamentally broken.

For decades, we have been trapped in a false binary: bootstrap (and struggle for years) or raise VC (and give up control).

But in 2025, AI has changed everything.

I'm seeing a revolution in how companies are built, and the smartest founders are leveraging a new emerging model that almost nobody is talking about.

Here's the insider view of all four approaches to building and funding a company, with hard numbers from the trenches. I have gathered them after chatting with 100+ founders and learning from the smart founders on the Lean AI leaderboard:

The Traditional Options (That Are Failing Most Founders)

Option 1: Bootstrapping

You fund everything yourself, maxing out credit cards and emptying savings accounts.

You retain 100% ownership but face a lot of challenges:

90% of startups fail within the first 3 years, and the failure rate is even higher for bootstrapped companies. 8 out of 10 bootstrapped companies fail within 18 months due to cash constraints.

Your personal finances bleed red for years, with no guarantee of survival.

Even successful bootstrappers often take 5+ years to reach mere six-figure incomes (that too after working 80-hour weeks for less than minimum wage).

Option 2: Venture Capital

75% of VC-backed startups never return capital to investors.

Only 0.1% achieve the unicorn exits we read about in TechCrunch.

Yet the model forces ALL founders to operate as if they will be in that 0.1%.

Founders surrender chunks of equity at every round: 20% at Seed, 20% at Series A, 15-20% at Series B...and so on.

By Series C, founders typically own just 15% of their companies, and 99% never reach that milestone.³

A founder who builds a $50M company with VC often walks away with less personal wealth than one who builds a $10M company via bootstrapping.

Option 3: Boot-Scaling

You self-fund until you demonstrate meaningful traction, then take a massive funding round (often from private equity).

This approach preserves early ownership but comes with hidden costs:

You endure the cash-strapped struggle of bootstrapping for years.

Then, you dilute massively in a single event (often 40-50% of your company). You also lose control to micromanaging private equity buyers, who often ruin the company culture.

The risk profile is high: You burn personal runway and then bet everything on a single scaling event that fails 72% of the time.⁴

Subscribe now

The New Model (That AI Has Made Possible)

Read more

Image source: see the work of the illustrator.

Hey Everyone,

In this article I want to cover some of the reports I’ve been reading, namely the Stanford State of AI Report 2025. Full disclosure you might want to print this one out, it’s infographics utopia. Whether you are an investor, China watcher, professor, analyst or someone in industry, we’re going to cover some major AI trends and emerging tech snapshots here.

I’m going to make your job more easy if I can. I will link to the original reports repeatedly throughout the article.

Browse the Google Drive

Meanwhile Trump’s tariffs, especially the China trade war now escalated could hurt AI and datacenter by making stuff more expensive, lowering BigTech margins and disrupting critical supply-chains for advanced technologies.

In the open-source AI space, Together AI, Alibaba and Deep Cogito are doing some interesting things.

Share

State of Open-Source LLMs

While Meta’s Llama-4 appears to be a disappointment, and we usually think of Mistral, DeepSeek or Qwen in Open-source LLMs, I want to turn your attention to a couple of other contenders (though as we will see they are related) I think deserve a worthy mention.

Together AI who raised over $300 million Series B a month ago, have announced 🧠💻 DeepCoder-14B – A fully open-source, RL-trained code model! It’s interesting because it’s a code reasoning model finetuned from Deepseek-R1-Distilled-Qwen-14B via distributed RL.

Read blog

Remember this is open-source, they have basically democratized the recipe for training a small model into a strong competitive coder—on-par with o3-mini—using reinforcement learning.

Qwen 3 Upgrades from Alibaba Cloud

The next version of the Qwen LLM series, Qwen3, brings a new level of advancement in both natural language processing and multimodal capabilities.

• Qwen3-VL (Vision-Language) is a powerful addition to the Qwen3 series, designed to seamlessly integrate visual and textual data, enabling the model to understand and generate language in the context of images and videos. With advanced multimodal capabilities.

• A significant highlight of Qwen3 is the introduction of Qwen3-math, a model designed specifically for tackling mathematical reasoning and problem-solving tasks.

• In addition to these enhancements, Qwen3-Audio makes its debut, bringing robust audio processing capabilities into the mix.

We will know more about Qwen3 when they announce the flagship model which is expected soon.

What is Deep Cogito?

Coming out of stealth this week, Deep Cognito introduces revolutionary open-source hybrid models. They boast a family of openly available AI models that can be switched between “reasoning” and non-reasoning modes. As you know, Hybrid models can quickly answer simple questions while spending additional time considering more challenging queries.

Read the Blog

• They are releasing open models of 3B, 8B, 14B, 32B, and 70B sizes.

• They plan to release larger models too including 109B, 400B, 671B, in the coming weeks / months, as well as improved checkpoints for each of these model sizes.

• The LLMs are trained using IDA - a scalable alignment strategy for better performance, according to the CEO.

• You can download them on Hugging Face or Ollama, or use them as an API on Fireworks AI or Together AI.

Deep Cogito says Iterated Distillation and Amplification (IDA) is a promising alignment technique for self-improvement. The Open-source LLM space is so much more collaborative you’ll notice and the team thanks Llama Team, Qwen Team, HuggingFace, RunPod, Fireworks AI, Together AI, Ollama, Nebius.

How the Media is Framing the HAI Report

The AI Index report tracks, collates, distills, and visualizes data related to artificial intelligence (AI).

Multiple Media outlets have taken the main talking point from the State of AI report that China is catching up in AI. The U.S. vs. China AI race for AI Supremacy is the main thesis of this Newsletter. So in this post we’ll be exploring this from multiple angles.

Many leaders in Silicon Valley and D.C., including in the Trump administration, say winning this AI competition is critical to the future of U.S. national security. But two reports this week highlight China’s gaining superiority. This one and this one. What should we make of this relative to the absurdly high trade tariffs Trump has put on China?

“The tariffs will make building AI datacenters much more expensive, both because AI servers are largely imported and will face tariffs, at least until supply chains can be rejigged, and because much of the other equipment in datacenters, like the cooling and power infrastructure, is imported as well,” says Chris Miller, author of Chip War.

The ASPI report is even more urgent. In a new report, the Australian Strategic Policy Institute (which I will refer as ASPI in the rest of this article) says China is outpacing the US in 37 out of 44 critical technology research areas. I’ve been reading this report for years and it’s moving in the direction of Chinese AI Supremacy. Yet hardly anyone in the West is talking openly about this! What it amounts to is China leading in the next era of emerging tech - from quantum to biotechnology, to space-technologies and robotics.

• Robotics

• Biotechnology

• Space-Technology

• Quantum computing

• Artificial Intelligence

• New energy systems

• Renewable energy

• Infrastructure construction

• Advanced materials & manufacturing

While the U.S. is hype driven around BigTech profits related to Generative AI, China appears to be taking a more holistic approach to technological innovation focusing on shoring up its semiconductor industry, winning the humanoid robotics race and dominant innovation at the bleeding edge of many frontiers.

Australian Think Tank’s Dire Warning

The Australian Strategic Policy Institute (ASPI) is an independent, non-partisan think tank that provides analysis and advice on defense, strategic policy, and national security issues, aiming to inform public debate and influence government decision-making.

• According to this Australian Strategic Policy Institute: China’s global lead extends to 37 out of 44 technologies that ASPI is now tracking, covering a range of crucial technology fields spanning defence, space, robotics, energy, the environment, biotechnology, artificial intelligence (AI), advanced materials and key quantum technology areas

While the Trump Administration is supposedly full of China hawks, he’s increasing tariffs and export controls, basically pushing China to innovate harder and more trade relationships with the rest of the world more aggressively. He’s boasting about investments in the U.S. which are exaggerated don’t address the key technologies of the future.

“China’s overall research lead, and its dominant concentration of expertise across a range of strategic sectors, has short and long term implications for democratic nations. In the long term, China’s leading research position means that it has set itself up to excel not just in current technological development in almost all sectors, but in future technologies that don’t yet exist.” - ASPI’s verdict is pretty stark.

A report of China using Quantum computers to fine-tune LLMs really caught my attention. It hasn’t been independently verified though or even peer reviewed.

The Stanford Institute for Human-Centered Artificial Intelligence (HAI) is perhaps one of the more neutral AI reports out there. Yet their data on China in AI says the same thing.

Packed with new data and in-depth analyses, our latest report highlights the most critical trends in AI – from shifting geopolitical landscape and rapid technological evolution, to AI’s expanding role in science and medicine, business, and public life.

Read the Report

• This is the report that we’ll be going into the most detail today.

• The ASPI report highlights a lot of important topics for our emerging tech coverage however.

• Also worth checking out: ASPI’s new Critical Technology Tracker.

  https://techtracker.aspi.org.au/

• You can compare the U.S. with China directly on any of their categories directly:

The results are a bit humbling. Their report is 84 pages.

Is China the Future Emerging Tech Super-Power of the 2030s?

“Our research reveals that China has built the foundations to position itself as the world’s leading science and technology superpower, by establishing a sometimes stunning lead in high-impact research across the majority of critical and emerging technology domains.”

All signs point to China really gaining the lead in the 2030s. It must overcome some pretty significant barriers in the semiconductor industry, the Cloud, its Venture Capital system and the way it develops startups, and its lack of foreign investment since the U.S. has been working to undermine this, along with export controls on AI chips. However nothing the U.S. is doing is likely to prevent the outcome that China will dominate, in most technology sectors.

I mean, I could make the argument China already has the lead in humanoid robotics and Open-source LLMs already in 2025. The U.S. inspite of the hype around OpenAI and Anthropic’s Generative AI capabilities, appear to be losing momentum in most of these categories to China in the mid to late 2020s.

Indeed even the way the Trump Administration interprets trade with China seems to be all wrong. Cleary the U.S. is not planning for a future where it is competitive with China in innovation and emerging technologies with these weird trade war last ditch efforts slow them down type policies with cumulative tariffs of at least 104%. If anything, the U.S. is too defensive and thinking about deterrence, insteading of promoting science and technology at home. Instead of building AI Infrastructure in one concentrated area, it’s slipping behind in many of the others.

If Jensen Huang is the New King of Tech? What is China? Nvidia’s stock remains down 30% this year and likely going down further. While Western media keeps talking about Silicon Valley or Elon Musk, the world is moving on. The Billions the Americans are spending, aren’t even in critical or right areas to keep up. The U.S. is cash rich, but not spending in the right places.

According to Stanford, the U.S. saw $109.1 billion invested last year —nearly 12 times China's $9.3 billion and 24 times the U.K.'s $4.5 billion.

ASPI analyzed the top 10% most highly cited research publications from the past five years on each of the 44 technologies to draw some of its conclusions.

Meanwhile in AI, China continues to lead in AI publications and patents. The HAI’s report is 456 pages.

China will be the Space-Tech Leader

Where China’s lead is likely to be the most impactful in the next decade is the defense and space sector. A key area in which China excels is defence- and space-related technologies. China’s strides in nuclear-capable hypersonic missiles reportedly took US intelligence by surprise in August 2021.

China becoming the space-tech leader will have ramification in national defense and especially future energy systems. Innovation as China is well aware of isn’t just about AI or generating hype for Generative AI that isn’t even the right architecture for major progress to occur. If Silicon Valley operates on hype and moonshots, China is a bit more grounded in how it invests in the long-term future.

At a certain point, American BigTech will just be left behind as they are revenue and shareholder driven (short-term), not innovation driven. Outside of BigTech what does the United States even have but a world leading education system and immigration system to funnel talent to their same companies? America’s vast Venture Capital apparatus isn’t even investing in the right things to keep up.

The main conclusion from the ASPI report that syncs with my own assessment is simple: China is further ahead in more areas than has been realised. The U.S. media seem to not want to cover this reality. Thus you can see more Substack Newsletters delving into it instead as independent curators, researchers, analysts, academics and journalists.

Which other Countries are in the Mix in Emerging Technologies?

ASPI’s global assessment of technology is fairly well done: so Outside of China and the U.S. we can see as follows:

“The race to be the next most important technological powerhouse is a close one between the UK and India, both of which claim a place in the top five countries in 29 of the 44 technologies. South Korea and Germany follow closely behind, appearing in the top five countries in 20 and 17 technologies, respectively. Australia is in the top five for nine technologies, followed closely by Italy (seven technologies), Iran (six), Japan (four) and Canada (four). Russia, Singapore, Saudi Arabia, France, Malaysia and the Netherlands are in the top five for one or two technologies.”

1. China

2. United States

3. United Kingdom

4. India

5. South Korea

6. Germany

7. Australia

8. Italy

9. Iran

Now we turn to the main strength of the ASPI report: (easy conclusions from the analysis)

Technology Monopoly Risk Chart

ASPI really breaks it down for the U.S. vs. China race:

You will notice China’s lead in many areas means it cannot really lose the robotics, Space-technology, Space-construction and hi-tech races of the sci-fi future. Or even a prolonged hot military confrontation due to its manufacturing prowess. Just don’t expect Trump to show weakness even though this is obvious to National defense types and think tanks.

What becomes absurdly noticeable is how few of these categories the U.S. is leading according to ASPI. That is a fairly poor sign of America’s ability to compete in the 2030s in emerging technologies if we are looking to the future of technological supremacy by nation-state.

Advanced Materials and Manufacturing

Artificial Intelligence and Communications

Subscribe now

Read more

Welcome Back,

There are days where I feel like a business publication and not just a technology publication, this has to be one of those days. Some stories are too big not to tell.

The dot com bubble burst 25 years ago to the day. On March 10, 2000, the Nasdaq Composite index peaked at 5,048.62, more than double its value from just a year earlier. It’s hard to imagine, but eventually, the Nasdaq shed more than 75% of its value, wiping out $5 trillion in market capitalization.

In the Generative AI era, Nvidia’s stock has been seen as the bellwether of the movement since around 2022. Here we are suddenly in 2025 and what’s going on with Nvidia? I ask because it’s lost over $1 Trillion in market cap in just a few weeks. Nvidia is down over 30% from its highs. It’s not many days where the NASDAQ 100 is flirting with being down 3% in a single day, so how about nearly 4% (3.81%)? March 10th, 2025 will go down as one of these rare days.

Nvidia is itself down 22.65% year to date on the stock market, likely closer to 25% after today. But who’s counting? Semiconductor companies are at the center of the AI boom and in freefall with Trump acting as an agent for chaos and confusion where economic uncertainty anxiety is flirting with recession fears. Nvidia as an AI stock is caught up in all of this.

Economists say Trump is proving to be an “agent of chaos” with his unpredictable trade tariff policies, but recession fears are overblown. Goldman Sachs recently cut its economic growth outlook due to the potential impact of Trump’s tariff policy and without a Fed put or a “Trump put”, there was a capitulation in the markets that was nearly unimaginable even a few short weeks ago. With tariffs looming, this could last a while.

Nvidia’s stock was down in a big way for the day on Monday March 10th, 2025, to the tune of 5% in a single trading day. There are many layers to Nvidia’s poor stock performance in 2025 and we need to peel the orange back to understand it better. 🍊 This is a day when understanding the stock market is part of how we understand AI. So buckle up if the topic interests you.

I turn to my Nvidia expert for a quick take. I offer more color to his points after the guest contribution.

Nvidia Is Having A Rough Start to 2025. It's Not Just DeepSeek

“Nothing lasts forever” is what Nvidia’s $NVDA shareholders may be thinking as the stock fails to see any revival this year.

Nvidia’s stock is possibly seeing its worst start to the calendar year in over a decade as a range of headwinds keep forming dark clouds over the GPU maker’s outlook this year.

DeepSeek, chip export regulations, scrutiny over sales of its accelerator chips or GPUs, and competitive pressures have all converged together, digging its heels firmly in the sand and draining all the exuberance out of Nvidia’s stock.

Of course, Nvidia’s bulls will laugh off any mention of competitive pressures that the stock sees.

But that still doesn’t justify why Nvidia’s investors have scaled back their investments in the stock with the company losing $870 billion in market capitalization from Nvidia’s market cap peak in November. This is a stark reversal of fortunes for Nvidia, its investors, admirers, and its rockstar CEO, Jensen Huang, who, at one of the many peaks of Nvidia’s exuberance last year, was signing t-shirt(s) for his fans.

Nvidia’s headwinds show no sign of dissipating yet, especially DeepSeek, which emerged as one of the earliest, deadliest threats to Nvidia this year.

This Is How DeepSeek Sparked Nvidia’s Selloff

Every AI enthusiast knows that the fierce competition for dominance in GenAI has centered around the idea of “scaling laws.”

Large models have large volumes of model parameters and are typically trained on larger amounts of data, measured in tokens, which often require copious amounts of compute resources (more computing chips and power), thus enabling the model to get more “capable,” a.k.a. perform more complex tasks and score better on benchmarks.

Training these models costs millions of dollars due to the vast scale of GPUs needed to train models. That’s why Nvidia benefited because every AI company was rushing to make their AI models more “capable.”

But when OpenAI launched the o1 reasoning model last year, a paradigm shift occurred in the narrative with experts debating that scaling laws don't scale enough.

Recent works

1. AI Networking Companies Are Set For Capex Primetime Mar ‘25

2. Nvidia’s Q4 Earnings Review: Chip Wars Are Heating Up Feb ‘25

3. How Nvidia Gets Impacted By The AI Chips Ban Jan ‘25

4. The One Hyperscaler That Stands Out Among Rising AI Capex Concerns Feb ‘25

🟠 Check out the Pragmatic Optimist to receive “alpha generating investment ideas” in the AI value chain.

Go Alpha 🗿

• For less than $2 a week, get my very best coverage and those of the emerging writers that go deeper.

Support this publication

Read more

Good Morning,

Today we are going to talk about reasoning models and their capabilities. Anthropic the makers of Claude, have released this week a new ‘hybrid reasoning’ AI model. While OpenAI and oth…

Read more