Token capital. Queryable institutional memory. Private traces from inside the org. A loop that compounds and stays yours across any model.
Satya just wrote about what we are building at @leveragecpu for quite some time now...
Slack was designed for humans sending messages to humans.
The next workspace has to assume humans and agents work in the same place, with the same context, under the same permissions.
That changes the product surface. It also changes the infrastructure underneath it.
Idk about looksmaxxing, but human <> human is pretty obviously getting more valuable atm, not less. Everyone selling "fully autonomous business in a box" is racing each other to zero. Once anyone can spin one up, the business is worth what it costs to copy, which is basically nothing. And the value moves to whatever stays scarce. and the one input you can't synthesize is a verified human on the other end. opposite of artificial is natural
Tyler Cowen's recent talk on what will matter post AGI was really good.
Buy: embodiment, physicality, charisma, looksmaxxing, taking the initiative
Sell: writing, being a managerial elite, universities
this is exactly what we’re building at @leveragecpu
a multiplayer AI computer where company communication, files, code changes, agent runs, tool calls, decisions, and memory live in one governed workspace
the company should be readable and executable by humans + agents at the same time
Increasingly, I believe companies may need to be rebuilt from the ground up, where you have a single timeline of all observability + product metrics + file changes laid out in a retrievable system, like Datadog + Posthog + Google Drive + Slack (really unified filesystem of Claude
Increasingly, I believe companies may need to be rebuilt from the ground up, where you have a single timeline of all observability + product metrics + file changes laid out in a retrievable system, like Datadog + Posthog + Google Drive + Slack (really unified filesystem of Claude
i am having more fun than ever collaborating with people
get an agent in the mix, we all bounce ideas off each other. prompt it, get more ideas. narrow in on something
stop dooming about ai and your job and start gangprompting
if you’re still writing loops that prompt coding agents you’re falling behind. you need to build a meta agent that infers what loops you would have wanted based on your vibe and then write those loops
I want some kind of LLM workflow tool.
• Ability to manage a set of input files (Markdown or similar), plus other general-purpose context.
• With real-time collaboration. (And maybe some concept of snapshots or VCS integration.)
• And the ability to create/manage a inference
billion-dollar company is the Vercel for internal agents.
an employee writes a Claude Code agent, hits deploy, and gets hosting plus a Convex database in one click.
IT scopes what it can touch, watches every execution, and tracks token spend per agent
One small @leveragecpu workflow I like a lot:
You make a channel for interesting links. Stuff like articles, tweets, customer examples, competitor launches, whatever is relevant to the business.
Then an automation watches the channel and periodically turns the best stuff into content drafts, saves them in Google Docs, and links back to the agent session so the team can see how it got there.
It sounds stupidly simple, but this is the kind of multiplayer AI workflow that actually feels useful to many people.
Just drop things where the team already talks, and let the workspace turn shared context into work.
Great work by @Vtrivedy10@nikogrupen et al - great to see these results in Law, mirroring our experiments published yesterday in Medicine
1. Batch grading reduces cost by ~1 OOM
2. Small models reduce cost by ~2 OOM
In non-verifiable RL, where judge latency blocks samples reaching the trainer, judge selection is a crucial knob for training efficiency.
Medicine:
x.com/bertgodel/stat…
Can we design legal agent verifiers that are up to 1,000x cheaper?
Verifiers are LLM judges that check an agent’s work against rubric criteria: they're used both in agent benchmarking and as reward signal in post-training.
But verifiers can be a bottleneck at scale.
For
453 Followers 2K FollowingAutomation CPA | Exploring AI Agents | @HelloRentSmart | prev:@deloitte @ey_us I tweet about random things, but mainly tech and music
116K Followers 12K FollowingCEO @heyJasperAI, Fmr President @Dropbox, Founder of Socialcast (acq by VMware), VP Prod&Eng @VMware, Founded https://t.co/Y7bgHZYSOp (acq by AOL) & built lots for @Interscope
3K Followers 7K FollowingContrarian investor in the public markets. Not investment advice or a solitication to buy/sell securities. May hold/trade securities mentioned without notice.
793 Followers 3K FollowingProduct Design and UX/UI specialist. Leveraging my experience as a 2x Founder to build and launch successful products & websites. Also a Cracked Engineer
7 Followers 71 FollowingBeyond Video Games Art,
Artistic journeys through persistent worlds. ! 🎨🎮
Voyage artistiques a travers les mondes persistants 🎨🎮
10K Followers 3K FollowingBuilding RUVs and more for startups @rollupshq. Small angel investor. Dad. Prev: 2x founder and prior Special Forces medic (18-D).
1K Followers 2K Followingmember of capital allocation staff @scalevp | previously ML, product, and engineering | math & cs @columbia | I ♥️ SF & bay area sports
41K Followers 1K FollowingStaff Research Engineer at @GitHubNext. But actually a designer. Writes about design, dev, & anthropology.
Also on Bluesky 🦋 @maggieappleton.com
32K Followers 4K Followingtweets about AI and other fun stuff. currently @foundationcap; wrote the context graph paper.
previously McKinsey, @georgiatech, @stackfolio (acquired),
66K Followers 2K FollowingChief of Staff at Remilia Corporation @remiliacorp333
Warlord Commander at YAYO Corporation @YayoCorp
THIS IS NOT A PROMISE OF EQUITY OR OWNERSHIP IN ANYTHING
804K Followers 967 FollowingTechnology's daily show. Hosted by @johncoogan & @jordihays. Streaming live 11a-2p PT every weekday. Sign up for TBPN's daily newsletter at https://t.co/Nhf5ohjInO.
65K Followers 59 FollowingStudent of mind and nature, libertarian, chess player, cancer survivor. @ Keen, UAlberta, Amii, https://t.co/u8za2Kod54, The Royal Society, Turing Award
3K Followers 1K FollowingDesigning and building. Currently autonomous firefighting drones @seneca_systems. Before design staff @perplexity and spacecraft operations interfaces @NASAJPL