That's what everybody in America told me: "I love to drive. I would never let a computer drive." Even in San Francisco, people would tell me that. Yet Waymo numbers are going up, and Tesla numbers are going up. Why? Because driving really sucks. When you get an FSD car, you're
MARKET TRENDS
-

AI News Roundup Covers Scientific Results, Agents and Terra Efficiency
By
–
Top stories in AI today: – Anthropic officially files to go public
– OpenAI’s next batch of AI science discoveries arrives
– Turn sales call notes into follow-up emails with AI
– OpenAI’s new Terra cuts the cost of AI computer use -
Daily AI Brief Covers Terra Computer Use and OpenAI DevDay
By
–
DAILY AI BRIEF 🗞 — Oct 27 OPENAI 🔥:
– GPT-6.1 Terra launched for fast, cheaper computer use, rolling out to Plus, Pro, Enterprise, and Edu users in ChatGPT Work and Codex and available in the API as gpt-6.1-terra at 50% lower prices than GPT-5.6 Terra.
– OpenAI held DevDay -

Commentary Examines Compute Providers’ Economics and AI Infrastructure Demand
By
–



The $3.1B/month rate isn't guaranteed through 2029. Regardless, the compute play is paying off for SpaceX(AI)>(SI). While neoclouds might suffer, the big compute providers will absolutely dominate. With NVIDIA being amongst the big winners. Buy more, save more.
-
Discussion Describes Competition and Collaboration Among AI Model Developers
By
–
There are currently five major and many small AI companies in the US that fiercely compete about building and serving the best models. Open source models are not far behind. There is massive collaboration with academic labs too.
-
Commentary Examines How Inference Revenue Relates to AI Research Costs
By
–
At the moment, the large AI companies are extremely benevolent. Their inference business is not sustaining the training and serve as demonstrations and PR, and their scientific efforts are basically meant to be a gift to humanity, because there is no business case behind them.
-
Commentary Compares Opus 5.5 With Recent Competing Model Releases
By
–
34 days.
— God of Prompt (@godofprompt) 27 octobre 2026
That’s how long Opus 5.5 has been the model everyone else is chasing. In AI time, that’s a decade.
Since September 22, we’ve had Kimi K3.1 Pro, DeepSeek V4.1 Flash, Nanbeige 4.3, and a pile of benchmark wins.
The benchmarks moved. The crown didn’t.
Opus 5.5 is… pic.twitter.com/JcaWSpr8Oq34 days. That’s how long Opus 5.5 has been the model everyone else is chasing. In AI time, that’s a decade. Since September 22, we’ve had Kimi K3.1 Pro, DeepSeek V4.1 Flash, Nanbeige 4.3, and a pile of benchmark wins. The benchmarks moved. The crown didn’t. Opus 5.5 is
-

Estimate Separates Anthropic’s Internal Compute Spending From Customer Inference
By
–

We can estimate the amount of compute Anthropic used for R&D/Training/Internal use vs serving customers – by looking at their compute spend (c.$9b in H1 2026) cost of revenue ($2.4b) which leaves us with the rest ($6.6b) – so around 70% of compute is internal spend. I wouldn't
-

Goldman Sachs Reports Links Between AI Adoption and Employment Growth
By
–
➡️ @GoldmanSachs “is starting to see clearer signs that AI’s impact is spreading beyond the tech sector”; “the share of US companies that have adopted it in their routine activities now shows a statistically significant negative correlation with employment growth” [Traduit de FR vers EN]
-
Anthropic Filing Shows Lower Compute Costs Relative to Revenue
By
–
My initial S-1 thoughts: As I expected, Anthropic is still firmly unprofitable. But, to their credit, they've significantly reduced compute costs as a % of revenue, leading to much narrower losses. On first look, there is a plausible and imminent path to profitability. This is