The labor-displacement framing tends to land as 'AI does the task, workers get laid off,' but the near-term effect that's actually showing up is internal stratification within companies before any headcount change. Workers who can run multi-step agentic workflows (Cursor with parallel subagents, Codex with computer use, Claude with full Cowork stack) are out-producing colleagues at the same role and seniority by margins that used to require seniority gaps to explain. Companies notice but mostly don't act on it for 12-18 months because hiring decisions and performance reviews lag tooling shifts. So the suffering shows up first as 'I'm being outperformed by people in my own role at my own company' rather than 'I lost my job to AI.' That's a different policy and politics problem than the layoff framing. It's not unemployment. It's a class divide built on AI-tool access and the cognitive habits to use it well, sorting people inside the same nominal employment status. By the time benchmarks like the one you cited (Composer 2.5 vs Opus 4.7 vs GPT-5.5) start setting expectations for what an individual contributor is supposed to produce per week, the layoffs become the second-order event — the differentiation already happened in the per-task elasticity gap.
Ilya Sutskever testified Sam Altman constantly lied and pitted executives against each other. Altman testified Musk demanded 90% control. The OpenAI trial is revealing ugly internal politics on both sides.
Honestly I found the trial super interesting and what it revealed that we already knew but with more context and details.
Brockman who has financial ties with Sam Altman has now been given control over the entire open AI Product.
I think their ambition to merge Codex with the ChatGPT interface is historically bad. The company's struggles are a textbook case of poor product vision and implementation. All the capital in the world won't save you if you have poor leadership. And I expect OAI to continue to struggle against competitors.
Sam Altman is it talented angel investor, he should focus on what he does best. A great CEO for a company going public? Not so much. Elon Musk and Altman are fundamentally past their prime and best years vying to remain relevant. I see them essentially as caricatures of the past. Their two IPOs are going to be heavily scrutinized both having somewhat dubious paths to profitability.
An administration of widespread corruption, inside favors, financial transactions and imperialistic pursuits destroying the credibility of America to its historical allies.
A small cluster of firms captures the upside, models, compute, IPO wealth, enterprise workflows, defense contracts while the public absorbs the downside: job insecurity, higher power costs, local data-center burden, surveillance, and institutional dependence.
The AI boom has become political, in a world of critically poor leadership and authoritarian decision making. When the affordability crisis is ignored so elites can profit more, the imbalances in the system are magnified.
Concentration risk of BigAI can lead to several unfortunate outcomes for civilization. There's no doubt that the United States is trying to leverage its AI Supremacy in nefarious ways. It's already crippling the rest of the world like with the Iran crisis and creating a geopolitics of conflict that's good for American weapons sales. Mythos in the Pentagon is just the beginning.
The demand for compute is creating scenarios for high-powered partnerships that belong to another era.
Google and Blackrock are teaming up to build a Neocloud with Google's TPUs. A reality where capacity is permanently constrained and semiconductor supply chains are continually behind demands.
The labor-displacement framing tends to land as 'AI does the task, workers get laid off,' but the near-term effect that's actually showing up is internal stratification within companies before any headcount change. Workers who can run multi-step agentic workflows (Cursor with parallel subagents, Codex with computer use, Claude with full Cowork stack) are out-producing colleagues at the same role and seniority by margins that used to require seniority gaps to explain. Companies notice but mostly don't act on it for 12-18 months because hiring decisions and performance reviews lag tooling shifts. So the suffering shows up first as 'I'm being outperformed by people in my own role at my own company' rather than 'I lost my job to AI.' That's a different policy and politics problem than the layoff framing. It's not unemployment. It's a class divide built on AI-tool access and the cognitive habits to use it well, sorting people inside the same nominal employment status. By the time benchmarks like the one you cited (Composer 2.5 vs Opus 4.7 vs GPT-5.5) start setting expectations for what an individual contributor is supposed to produce per week, the layoffs become the second-order event — the differentiation already happened in the per-task elasticity gap.
Ilya Sutskever testified Sam Altman constantly lied and pitted executives against each other. Altman testified Musk demanded 90% control. The OpenAI trial is revealing ugly internal politics on both sides.
Honestly I found the trial super interesting and what it revealed that we already knew but with more context and details.
Brockman who has financial ties with Sam Altman has now been given control over the entire open AI Product.
I think their ambition to merge Codex with the ChatGPT interface is historically bad. The company's struggles are a textbook case of poor product vision and implementation. All the capital in the world won't save you if you have poor leadership. And I expect OAI to continue to struggle against competitors.
Sam Altman is it talented angel investor, he should focus on what he does best. A great CEO for a company going public? Not so much. Elon Musk and Altman are fundamentally past their prime and best years vying to remain relevant. I see them essentially as caricatures of the past. Their two IPOs are going to be heavily scrutinized both having somewhat dubious paths to profitability.
Donald Trump's major big Tech trades while playing geopolitical kingpin is one of the most outrageous memories of 2026.
https://substack.com/@maverickequityresearch/note/c-260946426?r=cxsa3
An administration of widespread corruption, inside favors, financial transactions and imperialistic pursuits destroying the credibility of America to its historical allies.
The real AI backlash is not anti-technology.
It is a legitimacy crisis.
A small cluster of firms captures the upside, models, compute, IPO wealth, enterprise workflows, defense contracts while the public absorbs the downside: job insecurity, higher power costs, local data-center burden, surveillance, and institutional dependence.
That bargain will not hold.
The AI boom has become political, in a world of critically poor leadership and authoritarian decision making. When the affordability crisis is ignored so elites can profit more, the imbalances in the system are magnified.
Concentration risk of BigAI can lead to several unfortunate outcomes for civilization. There's no doubt that the United States is trying to leverage its AI Supremacy in nefarious ways. It's already crippling the rest of the world like with the Iran crisis and creating a geopolitics of conflict that's good for American weapons sales. Mythos in the Pentagon is just the beginning.
The demand for compute is creating scenarios for high-powered partnerships that belong to another era.
Google and Blackrock are teaming up to build a Neocloud with Google's TPUs. A reality where capacity is permanently constrained and semiconductor supply chains are continually behind demands.
https://www.businessinsider.com/smart-people-react-google-blackstone-collab-5-billion-ai-company-2026-5
For those like Michael Burry, the centralization in Venture Capital pouring into AI is historically concentrated: https://substack.com/@michaeljburry/note/c-261415340?r=cxsa3