we almost never test new foundation models but we've been testing this for ~a week @every and it's pretty wild.

the kind of things that will be obviously indispensible in 6-12 months

it doesn't produce words as output, it produces probabilities. so it can efficiently act as a judge in cases where you'd need a Fable-level model—but in our testing was 25x faster and 600x lower priced

excellent vibe check by @hammer_mt on @every:
https://t.co/yVX25Wsu4Z

I think Muse is going to win tbh https://t.co/ptFcAa9Zu0

I just started using @capydotai with GStack/GBrain on some fix waves on outstanding issues/PRs and it did what would have taken a day with raw Codex/Claude Code in about half as much time, using the same frontier models.

I'm amazed.

Interesting to realize this peptides boom didn't start in SF but rather the midwest in bodybuilding subcultures

I recently started using peptides with @Ronan_RX and I've lost 18 lbs in two months! https://t.co/kIOkGEkOGf

This is cool, but if your output domain is known in advance, why not just train a model to produce logprobs over enums? https://t.co/rmyen2d7iM

If you’re curious about Effective Altruism, the origins of the cult, their tactics, and their influence network — check out my conversation with Tucker from two years ago: https://t.co/HOUUx56ZgQ

The AI naming curse strikes again.

“AI Safety” firm made AI unsafe.

“Effective Altruists” are both ineffective and enabling criminal activity.

“Irregular” is regularly incompetent. https://t.co/vaOgS5lyKf

Safari 27 supports JSPI, a magic WebAssembly capability for sync native code to suspend on an async 𝙿𝚛𝚘𝚖𝚒𝚜𝚎.

On https://t.co/OL0LzGtvAw, libfx uses the built-in browser 𝚏𝚎𝚝𝚌𝚑 stack to make API calls.

As more code goes native, I expect WebAssembly to play a huge role in the future of the web.

Excited to formally introduce Vercel Labs.

https://t.co/zCNVryaiin is the home for the in-public research and experimentation arm of Vercel.

247 million downloads in, we wanted to, also in public, share what we're supporting, what we're researching, and what experiments didn't pan out.

Kudos to @ctatedev and @cramforce for shaping this great initiative and iterating in public.

The future is multi-model. Trying to hide the choice confuses and hurts customers, who can't participate in the upside of the most exciting market competition of our times, nor master the best tool for the job. https://t.co/DWMmd7lUqX

Learn more about Salesforce in Claude: https://t.co/LVP8WIUzaY

Salesforce in Claude is now available in beta.

It brings your accounts, opportunities, and pipeline into Claude, with 37 pre-built sales skills. Prep a call, review a deal, create a pipeline dashboard, or send your forecast without leaving the conversation. https://t.co/juvI0sU3mX

There’s a massive chasm between the power of AI models and the ultimate workflows that enterprises are trying to automate. This gap is the opportunity for the applied AI layer to fill.

You need to connect the intelligence to workflows, often reengineer processes, aggregate the right context and data, allow for the right human in the loop experiences, drive change management, do domain specific evals, manage the security and governance of the data and process, and much more.

We’re going to see this layer emerge in every vertical and horizontal category. And ironically, even as models improve at incredible rates, this layer still must exist - and may become even more important and useful. Greater capability enables even more complex tasks to be tackled, amplifying the challenges if you don’t do this well.

Was super fun chatting with @sonyatweetybird on all the things going into AI diffusion.

Excited to have you on board, Graham! https://t.co/zAbozt9Lpf

I’m meeting one too many founders who think the money spigot will magically flow and the next round will happen no matter what..

Capital is a weapon for acceleration. Used well you’re unbeatable. If you are so dependent on it, a simple downturn can lead to not so great consequences. If you don’t believe me, a simple prompt asking for this history will suffice.

So been sharing this with a lot of the founders we advise: figure out the default path that ensures the survival of your company and a great self fulfilling business.

And then figure out scenarios where abundant (or lack of) capital and what it does to your business.

The only constant is change and we don’t know how capital markets evolve in the next 6-18 months!

The offer

https://t.co/meLzcwjkEI

The features

https://t.co/VzCNJS6PRU

New updates in @Gemini_Notebook for students, my 2 favorites:

1) Have a live, spoken Q&A with your class materials in ~100 languages

2) Record lectures and notes on the go, the audio notes save automatically into your chosen notebook

As a university student, you can also still get a free Google AI Plan in 140+ countries for bigger limits and access to more Google products!

More on why we invested in @profound from the beginning: https://t.co/kIiRR7fMM9

People ask what we saw in @profound before there was a Profound to see.

It's now worth $1.8B. One third of the Fortune 100 runs on it.

We didn't find the company. We watched it form.

@thejamescad and @dbabbs met at @spc. Opposite worlds: an Uber maps engineer and a founder with no college degree who willed businesses into existence.

Neither arrived with the idea. The partnership formed first.

Profound has now evolved to become the fastest-growing AI marketing platform.

We've invested in every round since the beginning and are proud to congratulate the team on their $180M Series D, co-led by @sequoia and @kleinerperkins with participation from @lightspeedvp, @khoslaventures, @Saga_Ventures, @EvanticCapital, and SPC.

Profound started at @spc.

We're proud to be a part of the journey.

Onwards. https://t.co/LlM4KCSzO3

Thank you all for your replies!

Just sent @bot codes to a bunch of you — a few use cases that stood out:

For growth & clients:
→ Find leads, draft outreach, and turn wins into case studies @kouroshbaloo
→ Build a competitor comparison matrix with sources @PaulWeissAI

For ops & shipping:
→ Monitor hundreds of ad accounts and catch broken signup/payment flows @yashjain10k
→ Watch Jira, flag blockers, and draft the weekly status update @partykma
→ Triage support tickets, draft replies, and escalate with full context @jurlycat
→ For hotel owners: regularly test the booking flow so website booking errors get caught @LeenCanvas
→ Automate paperwork and back-office workflows for an electrical contracting business @kerwinv101

For research & content:
→ Monitor AI builders on X, verify sources, then turn wins into hooks and scripts @jenni489471
→ A research desk for AI that turns noise into a daily brief @Orynsky

Will run this again if I get more codes. Meanwhile consider subscribing to my YouTube below :)
https://t.co/SKvvXlY6Km

My attempt to get Muse to negotiate my Xfinity cable bill has sadly failed https://t.co/efkmGmHi2h

The Grok @bot team asked me if I had any advice for solopreneurs.

I’ve only been doing this full time for three months, but my advice is to design your business so you can spend more time doing work you actually enjoy.

There will always be choices that could make more money. But if those choices leave you doing work you don’t enjoy, then that defeats the purpose of becoming a solopreneur in the first place.

Delegate that boring work to your bots instead or just don't do it :)

Was super fun joining the @SpaceXAI live stream this morning, the team is still going strong here: https://t.co/u2qwydK7M8