'Gpt2-chatbot' Goes Live
'Gpt2-chatbot' Goes Live
Sam Altman is a sucker for publicity stunts. OpenAI is a parade of X campaigns. Is this GPT-4.5 Turbo Beta? What is SuperNova 5.0? Who are Chinese future contenders?
Hello Everyone,
Welcome back.
Something odd has been going on this week, the last week of April, 2024.
With OpenAI, there are so many new features, gimmicks and internet stunts we’re honestly getting a bit desensitized with it all. Especially on Twitter.
I was surprised there’s already been many articles written about this ‘Gpt-chatbot’ that feels a bit contrived.
📚 A Book I’d like to Recommend 🌟
is a prolific book author in machine learning, I highly recommend his work if you are learning about AI. “Machine Learning Q and AI”
30 Essential Questions and Answers on Machine Learning and AI. His book is a #1 New Release on Amazon.
Listen to my thoughts on this!
1 min 39 seconds:
For pure book authors in the machine learning space, I also can appreciate his Newsletter is free.
His Diagrams are very Clear
He has sent a copy of his new book to various AI authors including here
, another great thinker on AI here in Newsletter world. Two of the best!Nathan Lambert with his Copy
Reviews
“Sebastian has a gift for distilling complex, AI-related topics into practical takeaways that can be understood by anyone. His new book, Machine Learning Q and AI, is another great resource for AI practitioners of any level.”
, Writer of Deep (Learning) Focus
–
“Sebastian uniquely combines academic depth, engineering agility, and the ability to demystify complex ideas. He can go deep into any theoretical topics, experiment to validate new ideas, then explain them all to you in simple words. If you’re starting your journey into machine learning, Sebastian is your guide.”
– Chip Huyen, Author of Designing Machine Learning Systems
Back to our Topic
Back to the gp2 Twitter/X event that took place recently: I don’t always know what OpenAI’s comms teams are doing.
This is not so great PR
A so-called “powerful new AI system” has surfaced on the internet today (well the day before yesterday), sparking speculation about its origins and capabilities, with some researchers suggesting it represents a significant leap forward compared to existing AI models.
A lot of promotions are taking place on Twitter/X that are clearly not exactly organic. The gpt2 frenzy is a case in point:
The model, named "gpt2-chatbot," made its debut on LMSYS Chatbot Arena, a website known for comparing AI language systems, without any prior announcement or promotion.
Andrew Gao (see Tweet), a Stanford University student and AI researcher, tested the chatbot's mathematical abilities by presenting it with an International Math Olympiad problem. Impressively, the chatbot solved it on its first attempt.
According to Ethan Mollick (see Tweet), a professor at the Wharton School of the University of Pennsylvania who studies AI, in his experiments, the model outperformed GPT-4 on complex reasoning tasks, such as writing code to create a picture of a unicorn.
Meanwhile, AI influencer Rowan Cheung (see Tweet) praised the chatbot's proficiency in ASCII art, describing it as "miles ahead of any other model" in this area.
Furthermore, Chase (swee Tweet), a founding engineer at Codegen, stated that the chatbot's coding skills surpassed those of GPT-4 and Claude Opus. He mentioned that the gpt2-chatbot performed exceptionally well in complex code manipulation tasks, even outperforming newer models.
Have to say that it’s not clear if these are paid (Sponsored) posts by these AI influencers are legit honest reactions, which would be rather hard because they all occurred at around the same time, some 16 hours before I wrote this.
Rumors on X and x.AI’s Impressive Funding Round
Several users on X pointed out that the model identifies itself as "ChatGPT, a large language model trained by OpenAI, based on the GPT-4 architecture." Speculation that the model is the creation of OpenAI grew after Sam Altman, the CEO and co-founder of OpenAI, mentioned, "I do have a soft spot for gpt2" in a post on X. See Tweet. Now X isn’t good for much, but x.AI is due for some pretty serious funding. It’s been nearly a week since we found out that Elon Musk’s artificial intelligence startup X.AI Corp. is nearing a deal to raise $6 billion in a funding round that would value the company at $18 billion, according to a person familiar with the matter.
6 Bn. Round for ‘x.AI’ would be a Game Changer
Musk, a founding member of OpenAI, has had a contentious relationship with the startup in recent years. Meaning if he can actually raise $6 Billion that puts him with Anthropic as serious contenders competing with OpenAI directly and will improve their ability to hire the top AI talent (don’t call them “AI scientists”, they really are just engineers, while some are specialized in research).
However, it's still possible that "gpt2-chatbot" could have originated from a lesser-known company or research group aiming to showcase their AI capabilities and generate attention. Though I think that’s entirely slim to zero.
What is AI Anyways? Microsoft AI Chief Ted Talk
Copilot Workplace
Now with Github’s own Copilot Workplace, things are starting to get interesting!
Github is launching a a Copilot-native dev environment, designed for everyday tasks.
Emergence of “AI as a Service” is a Real Thing
You could argue that Microsoft is building a lot more utility into its products with Generative AI than OpenAI has been able to do itself, not to mention building more datacenters in far away places like Japan and Indonesia. Bill Gates certainly seems to be a believer. I can appreciate the Enterprise AI play for Cloud leaders to keep their growth a bit higher for longer.
Microsoft, Google Cloud and Amazon have done fairly well at this in the last 18 months using the Generative AI hype.
New Newsletter on Semiconductor Industry
I have a new premium Newsletter that will do deep dives on TSMC and related semiconductor stories many relating to Taiwan and the APAC region.
Learn more here.
Introducing: Semiconductor Reports
Semiconductor Things and Reports are my two dedicated Newsletters to learning more about the AI chip, semiconductor and datacenter and AI supercomputer space.
Semiconductor Reports ™ is for deep dives into the semiconductor and AI chip industry specializing in Taiwan affairs. It is the higher tier of my Newsletter Semiconductor Things.
GPT2 is all you need
Let’s dive into this gpt2 topic a bit further.
Let’s explore what SenseTime’s SenseNova 5.0 is and discover the other Chinese AI startups.
Keep reading with a 7-day free trial
Subscribe to AI Supremacy to keep reading this post and get 7 days of free access to the full post archives.