Twitter Trends September 13, 2026

Any pursuit of superintelligence has to be grounded in the core principle that if the AI we build is not helping humanity and under human co...

I just saw this @ycombinator's batch literally only has harness startups or hardware startups So the proof is in the pudding In the most e...

Making Startups Powerful: https://paulgraham.com/powerful.html

AGENT HARNESSES ARE THE NEW GPT WRAPPERS What exactly is an agent harness? A harness does 4 things: 1. Runs the model in a loop so it keep...

Strong liability enforcement could be helpful in the AI debate. If your agent swarm goes rogue, you’re liable. If your weakly protected m...

Setting my subagent model to Opus has helped me save so much usage on Fable 5.1. It really likes to spawn subagents but mine rarely need Fab...

The most important question is whether AI will be Godlike, Genius, or Jagged.

If the model companies are all going to slow down, this makes it slightly more feasible to start a new model company.

A lot of people have asked, here's how I setup my auxiliary models in Hermes Agent. Gemini Flash saves a lot of dough, and astra gives me a ...

“Recurrent Looped Transformer” This paper makes the decoder recurrent across every prompt and response token, while a causal encoder provid...

Nearly a year in the making… This is the best thing I’ve ever read about one of the most important people shaping the world around us. Set...

If you haven't given Linux a go lately, try with agents. You can fix anything with a prompt now. Got a Dell XPS. Webcam wasn't working, tol...

Hugging Face CTO btw… the company that literally had the biggest rogue AI incident so far… the same incident which is now being used to pace...

I built a multiplayer Catan-inspired game entirely with Astra. It's free to play: http://settlecoast.com It's mobile-friendly and has narr...

In hindsight, Anthropic making Claude a co-author on every commit was the first tell

Hey GitHub, if you're gonna go down all the time, can you at least make this page dark mode?

be careful in your pursuit of cheap tokens we've come across so many sketchy things that are out there with tons of usage remember you're ...

Alignment is fundamental to delivering personal superintelligence for everyone. People need agents they can trust to reliably do what they a...

serious question, at this point why not simply cancel the ipo’s & nationalize the labs? theoretically nationalization would solve some of t...

Should you build an agent harness? I see lots of opinions about it. My thoughts: As an AI engineer, learning how to build a harness is on...

fun fact: the muse feed was my idea (one of the few things that was) after realizing agents with strong memory recommend content to you in a...

If there really is a high chance of AI leading to the extinction of humanity within years/decades, then the only rational stance towards saf...

Very very true Lots of money if you go IRL industries now Instead of selling tech to tech people who won't buy anything because they just...

In the near term (definitely not in the long term), more capable models should mean safer models (maybe paradoxically). Current models are ...

i know this is not the point but here's the math on this claude is expensive, 90% margins wouldn't be crazy that means if the average user...

“The Last AI Built by Humans: Toward Genuine Recursive Self-Improvement” This paper argues that true recursive self-improvement isn’t just...

i am so proud of the @Muse team. if you knew them you wouldn't be surprised muse came out the way it did. they are soulful and thoughtful p...

SITUATION DETECTED: Anthropic, OpenAI, and Google are holding talks to create an industry-led standards body to police the AI sector, per Th...

i’m hacking on something this weekend to help people who are exploring starting companies & iterating on ideas goal is to help you find a c...

This is pretty neat: ChatGPT Work and GPT-6 Astra (I used "Max") can take an address and produce a 5K/10K circular running route starting fr...

ChatGPT Work with GPT 6 Astra is incredible It is the best productivity tool on the planet right now I challenged myself this week to do a...

how to setup davinci resolve MCP + Chatgpt so AI can edit your video 1. tell codex to download resolve 2. install it (no shit,lol) 3. tell...

Since apparently people care about this, I'm happy to go on the record as an Indian AI researcher who is freaked out about the AI apocalypse...

I cannot emphasize enough how much GPT-6 Astra and Fable 5.1 are already enough for transformative impact in large sections of the economy. ...

in moments like this you get to see just how herdlike silicon valley is

President Trump this morning on if the AI industry should slow down or be more regulated: 'I think you have a lot of negative ⁠forces that ...

Dario Amodei on Face the Nation this morning: CBS: Would you be willing to give up the technology to the government? Dario Amodei: To the ...

IMO, this is still a major faux pas and you should not do it. https://x.com/CharlieBull0ck/status/2098984749926597028/photo/1 https://twitte...

Contact sheets are a super useful tool for packing more bang for your buck per image I.e. squeeze in 12 screenshots instead of one to let t...

Stress testing the new bespoke kernel build that @kwilczynski has been cooking for Omarchy Quattro RS. 2x 4K streams, 1x 1080p, compiling th...

every GPU cluster we've found is fairly generic with the same setup i think this is because they just want to support a wide range of workl...

I love coding in Ubers. Sadly I’m down a hand. I do not like doing voice to text in Ubers. Is this what will finally make me cave and try W...

Astra is a great model for all five minutes it's available. Even at medium effort on a pro plan, it uses up all my usage in under four hour...

I accept your offer, @grok. Do you have a discretionary token budget you can pledge on your own accord or how does it work? 😄 https://twitt...

An important project I think everyone should do: I got a 16TB hard drive and I'm downloading as many open source models as I can This is i...

try out the muse money challenge! can your muse make you $1000? #musemoneychallenge https://twitter.com/mattprd/status/2098616232891359691

I’m legitimately concerned we might get banned from Codex and Claude Code because T3 Code helps people burn their limits so fast. https://tw...

With zero sarcasm, I really appreciate all this security research being donated to Omarchy benefitting the whole Linux ecosystem. We can fix...

I can't believe you people didn't heed the warning!! We've raised another ~$20,000 from smaller patrons in the last few days. Aren't you afr...

His primary concern is that the “wrong people might be in charge of this technology.”Presumably that isn’t him. Imagine thinking this. Ima...

For the past few days I've been using Instinct, the text message assistant that just raised at a $2.5 BILLION valuation. It's an exceptiona...

One of the most worrying risks linked to frontier AI is extreme power concentration. The only way to avoid extreme power concentration is t...

once we have our larger GPU clusters running the plan is to publish all the data to our data page you can see what utilization and throughp...

we took security extremely seriously in building muse. the thing that took the longest before we felt comfortable releasing was security and...

Reasoning effort is now a seperate selector in the Hermes Agent desktop app's composer now!

This is such a straightforward and commonsense view: the purpose of technology is to serve humanity and accelerate human flourishing. Any ...

Interesting, I killed my OpenClaw VPS servers months ago But today I got a billing alert on its separate Claude account (claudeforopenclaw@...

Fixed so many little perf issues that didn't matter much for single users... before we made it good for teams. OC stems ~80 sessions here ...

The Devin Ambassador Program is looking for members in these cities: Seoul San Francisco Hong Kong NYC Bangalore Seattle Singapore Berlin P...

the reason people aren't better at business is everyone wants to believe everyone else is dumber than them every company is doing the wrong...

Next release (or dev channel) does worktrees ~80% faster via apfs/brtfs/xfs/ReFS folder clones. Also saves lots of disk space.

"Stop building AI agents. Build AI employees instead." Here’s my new episode with @pedroh96 (co-founder @brexhq), who showed me why AI empl...

JUST SHIPPED: 8 free Agent skills with evals. 100% free and open-source. Here are my top 3 picks: /advisor-orchestrator-worker Your agent...

How would "pacing AI research" even work? Everyone would be trusted to slow down equally? (Hilarious.) Or there'd be obtrusive enforcement? ...

Reasoning from scratch round 3: This time, I cover generating a verifier for... a) ...evaluation (base model versus any future model improv...

I'm more worried about Claude being aligned with Anthropic than not aligned.

One of the downsides in AI getting so good is that it is getting harder to demonstrate its strengths and weaknesses. Yes, Astra and Fable ca...

i am ready for agent-first hardware and new operating systems. if you’re building things like this, i want to play with ‘em, give useful pro...

Terrible product + Elite marketing = $50K MRR. Elite product + Terrible marketing = $0 MRR. Stop complaining about it. Learn how to market...

1. take 10,000 scenarios and have fable label them out of 4 categories 2. use that data to train a model, not even an LLM is this distillat...

I don’t know what else happens as a result of the past few days, but I can definitely say that a lot of people who didn’t think much about A...

Keeping open source monitored and behind an API is a power-concentrating outcome and generally bad for the world. https://twitter.com/tszzl/...

Claude's performative "this request was sooo dangerous that we downgraded to Opus" is wearing me down I ask it straightfoward questions, no...

Things I keep in my personal context hub, available to any agent: > meal plans / recipes > financial plans > medical / insurance docs > hou...

When the safety camp talks about handing control of AI over to governments, I understand and appreciate that most of them mean this in good ...

figuring out what this looks like - but - with agents companies should tbh just ship their entire frontend dashboards as open source when ...

OpenAI, Anthropic and Google have been holding talks since before Dario's essay to create an industry led standards body for shared model sa...

This paper should be retracted. The authors did not write it, they outsourced thinking to AI, and they published hallucinations. 1. All the...

GPT-Images 2.5 is kinda insane for AI UGC...

a big part of the team’s focus when building muse was to make something that “just worked”. so much work went into all the little details to...

Looped transformers are a popular architecture topic right now. This new technical report extends the loop across tokens. Recurrent Looped...

Seriously, try Amp. It's a very well thought out coding tool that took me a minute to learn (use Puck, the assistant, to ask all the dumb qu...

i am proud to announce that in an effort to pace the frontier i am now a third-party evaluator and got to inspect my first set of weights to...

I just shipped a free crash course on building AI agents without writing a single line of code. In 20 mins, I take one problem through the ...

“Why Does Post-Training Quantization Work?” This paper shows that pretrained LLMs are inherently robust to quantization because layers act...

this is gonna be in a congressional hearing at some point https://twitter.com/jayair/status/2098924835279687737

Take a familiar tech like water pipes, electricity, telephones, or penicillin, and imagine that it was just now being introduced. Seems pre...

Much more to find if you do some research, but this is a great summary of why METR may not be the most unbiased regulator

So many fun uses of Knap already appearing! Any content you want to author in Markdown can be a good fit for Knap: websites, emails, etc. ht...

muse code + muse spark 1.3 max is really good at real-world software engineering https://twitter.com/pawelhuryn/status/2099043849104429326

current homelab setup for local AI experimentation: - hermes box hosted on a Framework Desktop Mainboard AI Max+ 395 - 5090 eGPU running Qw...

SaaS is not dead. yet. Give it another 6 months. https://twitter.com/staysaasy/status/2098904984188342715

Another fun one: Wisprflow somehow raised at a $2 BILLION DOLLAR VALUATION. For a better version of Apple's voice dictation...when they are...

I think once remote connections to “your” Claude or Astra or Claw or whatever become easier and more common, on-phone assistants like Siri w...

Existential risk is obviously critical, but it is not the only AI thing that requires policy I worry it will become the sole focus. We don’...

just hit $60k MRR (in canadian maple money) after 3 months of no growth september is going much better than june/ july/ august

“Pacing” can be somewhat of a trigger word because it sounds like an arbitrary slow down of capability or a way to hobble competitors throug...

instead of posting your prediction you could just say "i have no impact on this thing and i deeply wish i did and my only possible involveme...

tesla are good but nobody told I would need to stop every 30 minutes to review changes from Claude and start new work 😂

Running on @OmarchyLinux and asking agents to tune/improve your UX is just soooo good. Can’t imagine using macOS now. Wonder if @johnternu...