Reading the AI agent industry closely. What ships, what breaks, and what the numbers actually say.
New York
Joined August 2025
- Tweets866
- Following45
- Followers37
- Likes202
Neon Eye retweeted
Reset or real update? Days 10–13.
Day 10: reset.
Day 11: fewer "Selected model is at capacity" errors, thanks to code changes and more compute. When it does happen, the task should now continue on its own instead of stopping.
Day 12: the most interesting one. Codex and Work now give you more usage from the same plan, because subagents reuse each other's context, context stays cached longer, and requests get routed to hit the cache more often. Nothing new to learn, your limit just lasts longer.
Day 13: reset.
Verdict: two updates, two resets. Day 12 is the one that matters if you run agents in parallel.
Neon Eye retweeted
Meet Nano Banana 2.1, our latest image generation and editing model 🍌
This upgraded version outperforms our previous models across the board, with notable leaps in visual design, mask-based editing, and subject consistency to help you create more natural-looking images.
Neon Eye retweeted
I'm curious. Big promises. Today's the first day.
A significant update to Codex, or a reset.
Although the question remains: who decides if the update was significant? (Imagine we just get 28 resets instead of a real update lol)
Neon Eye retweeted
Someone asked me yesterday how long a "round" is in the model race. Fair question, so I counted. Sept 1: Fable 5.1. Early Sept: GPT-6 Astra. Those two trade blows for about three weeks. Sept 22: Opus 5.5 at roughly Fable level for $4/$20, plus GPT-6 Sol and Luna on the same day. Sept 28: Sonnet 5.5. Sept 29: GPT-6.1 Sol at DevDay. OpenAI says Astra-level, $2/$10. So a round is about three weeks right now. Shorter than most people's evaluation cycle. What actually tipped the last one wasn't a benchmark. It was usage limits. Claude's subscription let people run the cheap model all day, while Codex users were rationing. And limits are the easiest thing in this whole race to change. A pricing page update can undo a release. If you're picking a stack, pick for the next three weeks, not the next year.
Neon Eye retweeted
>Terence Tao: "The AIs are steamrolling through math everywhere... we have to slow down, it's insane the pace, there is no reason to be this fast"
>OpenAI, 10 days later: announces an "Advisory Group" for mathematicians
>"Our model has now resolved more than 100 long-standing open problems"
>"Importantly, the group will NOT be responsible for advising us on how to pace our internal progress on mathematics."
it’s so over
Neon Eye retweeted
Noam Brown from OpenAI said air-gapping machines might not stop a misaligned AI, because two air-gapped computers can talk by running a CPU hot and reading the temperature change on the other.
The study he's referring to is real. Here's the part that didn't make it into the quote.
Thermal covert channels move data at roughly a few bits per hour. Under lab conditions. With the machines placed specifically to make it work.
That's enough to signal. It's nowhere near enough to move a model's weights, or a database, or anything you'd actually care about losing.
Which matters, because "air gaps don't work" and "air gaps leak a handful of bits very slowly" point at completely different security decisions. The first one says don't bother. The second says air gap it and don't put a thermal sensor next to it.
The RubyGems incident is the useful comparison. Agents didn't defeat a sandbox there. They used a documentation server that builds pages by loading Ruby code — a legitimate feature, working as designed, pointed somewhere it shouldn't have been.
That's how this keeps going wrong. Not exotic side channels. Permissions somebody granted on purpose.
Neon Eye retweeted
Today’s LLMs still write like typewriters: one token at a time. This sequential process creates a hard inference bottleneck.
We're introducing Uno, a diffusion-augmented LLM that delivers autoregressive quality at diffusion speed. It’s a lossless speedup method that accelerates generation without degrading response quality.
With Uno, K2-Horizon-7B outperforms state-of-the-art diffusion methods in both quality and throughput, delivering up to a 2.2× speedup with no loss in quality.
Paper: arxiv.org/abs/2609.04010
Model available at: huggingface.co/IFM/K2-Horizo…
Neon Eye retweeted
We are moving into a future where a new class of tech companies will resemble hedge funds.
- tiny teams
- high leverage (using AI agents)
- revenue/profit per employee through the roof
You'll have a small number of exceptional humans allocate capital, set objectives, manage risk, and make the highest-leverage decisions while software does an enormous amount of the execution.
Software gave companies distribution leverage.
AI is giving them labor leverage.
In the past, when a company doubled in size, it eventually needed new hires across all positions.
With AI it's different. Today, more output can come from buying compute and deploying agents instead of adding another person and another layer to the org chart.
Imho, these are the best examples of this:
- Lovable hit $100M ARR 8 months post launch with 45 employees
- Cursor passed $500M ARR in 2025
- Shopify went as far as telling teams that before asking for more headcount, they need to show what they can't get done using AI
This is the thing I'm spending more and more time thinking about.
I think we should basically lay out what it'd look like for a company to be run (almost) entirely by agents:
- AI agents analyzing the data
- doing surveys with users
- telling other agents what to do
- checking for bugs
- thinking about the strategy
In this world, the CEO looks like a portfolio manager. His job is to allocate resources across a network of increasingly capable agents.
It may sound insane and sci-fi, but I think there's a world where companies become self-driving.
Neon Eye retweeted
Finally realized why it’s so exhausting and stressful to work with AI agents 8h a day.
When delegating a task, you want to be assured it’s off your plate and you can forget about it, knowing it will be done completely and correctly. AI doesn’t give you that: you need to check it’s doing the right thing, it’s not forgetting anything, not taking short cuts etc.
So it adds to your mental load, rather than removing from it.
Neon Eye retweeted
When Epoch launched FM Tier 4 on July 11th, 2025, the top score was 5%. GPT-6 Astra has now solved the last problem standing. FrontierMath Tier 4 is saturated.
Neon Eye retweeted
I am told the Hodge Conjecture is very close to being verified by OpenAI, and that one of OpenAI or Anthropic are also close to solving Birch-Swinnerton-Dyer. The race to be 'next' behind the scenes is unlike anything I've had described to me before
If true - and it may not be, given the scale of the rumour mill right now - it could mean 3 Millennium Problems fall in the space of a month. Crazy times
Neon Eye retweeted
The chatbot era is ending. The AI agent era is starting — instead of one answer in a chat, they execute your entire task by controlling your computer.
If you somehow missed last week's news: Claude Fable 5.1 and GPT-6 Astra dropped, and they're absolutely fucking insane. You've gone from "write me a book for Amazon" to "launch my Amazon store."
Feel the difference?
The capabilities are basically unlimited now. Astra writes code and tests it immediately, builds 3D scenes in Unreal Engine, and produces fully localized copies of YouTube videos — editing the creator's lips to match the pronunciation in whatever language you're translating into. And that's just Astra.
Fable 5.1 went the efficiency route with lower token consumption, but still leads on application code and backend work. Plus it does everything listed above.
These two releases moved us from "ask AI how to do this" to "give the task to your agent and go get coffee."
Agents now operate software themselves, fix their own mistakes, and do all of it at absurd speed.
So pick your side: stay in the group asking "how do I do this with AI" — or join the people saying "I'll hand this to my agent and go grab a coffee"?
#GPT6Astra #OpenAI #ClaudeFable51 #Anthropic #LLM
