In honor of reaching 45 million accounts on Bluesky, we will now shout out every Bluesky user individually 🧵 (1/45,000,000)

England needs to start thinking of itself as a hot country 🌡️

headless agents -> gradient-labs.ai/blog/how-and...

can't believe that Spotify hasn't built a "parenting" mode yet

"it's started rabbit holing" - we're using that to explain that weird state that LLMs sometimes get into where they start over-investigating instead of doing what you asked it to do

If you're in the camp where agent harness > LLM then the Fable news shouldn't be so important

Wild to read blog posts about Gradient Labs written by folks who aren't at Gradient Labs seriesastories.substack.com/p/the-sharpe...

2026 issue: my backlog of blog post drafts & ideas gets outdated if I let them sit there for a few weeks

Not having any EU competitive labs (plural intended) is about to get very painful.

Okay so Money 2020 is a pretty wild conference

Add to your watchlist if you’ve built a text agent and then need a voice agent youtu.be/8oNZ_vtoTb0?...

Ok this is a truly great tweet

This has devolved into

London, early June - Gradient Labs 🤝 Temporal meetup in our offices. luma.com/apqhg42n?tk=...

Who is proactively hiring software engineering interns these days, and what tips do you have for me?

The researcher vs engineer debate really encourages the member of technical staff job title

> Issue tracking is dead linear.app/next

It’s unreal being on this side of a LinkedIn automated outreach campaign

I don't expect it to convince anyone who isn't already convinced, but vintage language models solving coding problems is pretty good evidence of general reasoning ability

Accurate depiction of Claude & me www.reddit.com/r/VibeCodeDe...

In the last week I’ve heard of the “chat with AI” text box in products as both good (flexible to the long tail of user needs) and bad (you haven’t thought about user needs) Biggest product crisis since Google’s text box on a white page

I thought that in Engineering we had collectively moved beyond the “lines of code,” the “number of PRs,” and “tickets completed” metrics, but all of them are back in the spotlight now that AI writes the code. They are still as meaningless.

Richard Sutton talks about this

Everyone always talking about how awesome AI is. But have you ever tried keyhole surgery? 🤯

Hello there youtube.com/shorts/89h5U...

An unexpected valuable thing in b2b: people actually saying no when they aren't interested. Have started doing that as much as I can

Intentional or not, Principle Engineer (instead of Principal Engineer) is a brilliant job title

Even the models don’t like today’s state of flux

I have yet to be sent a link to a blue sky post, but still regularly get sent links from almost all other social networks. Enjoy

"""I had Claude Code implement a very simplified version of Git in 13 languages. Ruby, Python, and JavaScript were the fastest, cheapest, and most stable. Statically typed languages were 1.4–2.6× slower and more expensive.""" dev.to/mame/which-p...

Starting to think that the quiet foundation of tools that "increase productivity" is that they are addictive. Just like the way social networks grew with feeds. These tools are creating productivity by hitting us with some dopamine

Agents using code repos and ticketing systems is cool. Who is working on agent calendars so that I can schedule & see work that’s booked in?

We're currently comparing Claude & Codex. The most fascinating feedback I've heard about one of them is "for some reason I just hate talking to it" and I think that captures so much about what is difficult about building AI products these days

Someone I was chatting to at a conference said "AI should be the sat nav, not the car" to talk about AI adoption and so I asked him what he thinks about Waymo

Rapidly rebranding all my search benchmarks as eval awareness benchmarks

What I learned from being a CTO: there is always someone waiting for me to do something that I haven't done yet

Just over 50% of Gradient Labs is builders now, we wrote up a bit about our ways of working as we navigate this super competitive & ever-changing space blog.gradient-labs.ai/p/creating-a...

To date, HCI researchers have had no support on signaling their paper's relevance to AI, esp. when that connection is tenuous at best. We introduce a systematic framework to ensure LLMs are mentioned at every stage of paper reporting—from framing, to evaluation, to implications.

Vibe vs buy: you've never been paying for the code; you're paying for "someone else has thought this problem through in much more depth that I'm willing or able to"

it's been weeks and I haven't seen a single mention of the SaaS-pocalyse here

The idea that coding agents prove "learn to code was a mistake" is kind of analogous to thinking the printing press meant "psych! y'all wasted a lot of time learning how to write." Computational thinking is getting easier to pick up, and also providing more leverage than before.

Fascinated that my old modelstore library is still getting love (PRs, downloads), am throwing our friend Claude at some improvements github.com/operatorai/m...

We started working with Wise many months ago, am super glad that’s now public 🚀

We've launched our outbound agent today. It's been great seeing this come to life; an AI initiated conversation is so different from the reverse gradient-labs.ai/product/outb...

Cold reach out with: "I'd like to grab a coffee and understand where your business is headed" What do I even say to this

I mean if you care more about speed than fidelity, there are model choices you could make to speed things up... theonion.com/ai-chatbot-t...

Making AI agents _fast_ is more of an architecture problem than a model problem

"Who exactly is the intern?" benn.substack.com/p/the-gentle...