AI posts
People and companies shaping AI. Filter your watched accounts, catch up on recent posts, or find the most liked and viewed.
Compiled by the Toolstash team · Updated
209 watched accounts · 2977 matching posts · Showing up to 3,000 recent stored posts. Counts are snapshots, not live X totals. This page refreshes every minute while visible. Last collected .
@MiaAI_lab ·
Anthropic used an unreleased internal research model to find real zero-days in Chrome, Safari, Android, Windows, and Linux
62 CVEs patched in Chrome, all critical or high.
Source: https://www.anthropic.com/research/discovering-vulnerabilities-with-claude
@DeryaTR_ ·
I understand cancer deeply, down to the molecular level & I claim that AI will cure all cancer, along with every other disease, for that matter. And when that happens, this hematologist will be out of job, has a good reason to be worried!After all, no cancer, no cancer doctors!
@lmoroney ·
How much should you trust image leaderboards when you pick an editing tool? Microsoft AI's MAI-Image-2.7-Preview just landed at #3 overall in Arena's Image Edit Arena with 1470 points, a 45 point jump over MAI-Image-2.6, which sat at #7. It moved up in every category Arena tracks, including #3 for photorealistic and cinematic imagery, #3 for portraits and #4 for text rendering.
I'd use rankings like this to build a shortlist. Then take one real edit you do every week, like a background swap or fixing a sign in a shot, and run it through your top two picks side by side before you commit. 🎨
#AIart
Link in Comments
@theo ·
One of the cooler recent changes to T3 Code is the auto-promotion to the "best connection path" between machines.
Using a laptop to connect to your desktop in other room? Routes through your local network.
Falls back to Tailscale if available, T3 Connect if not
@ryancarson ·
Looking forward to recording this
@emollick ·
Important, but not unexpected, finding that experts working with AI agents are swayed by the positions of the AI, even if they think they are working independently.
If your independent judgement matters on a task, you need to be deliberate about when in the process to use AI.
@amasad ·
If I’m investing in frontier AI I would underwrite it mostly on new discovery and its applications like medicine because coding and most of knowledge work is quickly being saturated by sub-frontier models.
@AnthropicAI ·
We're expanding our partnership with @USGSA. Under a new OneGov agreement, we are bringing Claude to federal, state, local, tribal, and territorial governments at 40% off. https://www.anthropic.com/news/expanding-partnership-gsa
@hammer_mt ·
There has also been zero growth outside of big tech so that tracks
@poteto ·
i've gotten a lot of questions about how i manage my X account with grok bot, so i'm sharing my bot tibot today! he watches your mentions and:
• finds every post about you or any keywords, tagged or not • gives one short, sorted digest a day (or hourly, your call) • skips spam and noise • translates non-English posts • can auto block trolls and abuse • never posts for you
i still write all my posts and articles by hand but tibot has been very helpful with keeping me in the loop.
hire tibot today!
@ArtificialAnlys ·
Cartesia Managed Agents scores an Elo of 1115 on the Artificial Analysis Speech Agent Arena, 98 points above Cartesia's earlier voice agent product, Cartesia Line, with Task Success Rate up from 77.5% to 91.7%
Managed Agents is @CartesiaAI's fully managed platform for building and deploying voice agents. It is a cascaded system: Cartesia's own Ink Speech to Text and Sonic Text to Speech models run around an LLM of the developer's choice. We evaluated Cartesia Managed Agents with GPT-5.5 as the LLM and the Skylar and Daniel voices.
Key takeaways:
➤ Speech Agent Arena: Cartesia Managed Agents scores an Elo of 1115 (95% CI 1086-1144, 685 samples), 98 points above Cartesia Line at 1017 (95% CI 985-1049)
➤ Agentic tasks: On tasks that need tool calls, Cartesia Managed Agents scores an Elo of 1112
➤ Task Success Rate: 91.7% of eligible tool-calling conversations end with the correct final task-completing tool call
Congrats to the @CartesiaAI team!
See below for more details ⬇️
@alexolegimas ·
This is one of the most important pieces of social science on human and AI interaction that I’ve seen in some time. Experts follow agentic advice the vast majority of the time (causally), but only a tiny fraction credit the agent—majority think they came to decision on their own volition.
Reminds me of the classic Nisbett and Wilson study on people following outside influence without realizing it
“ But as more work is handed to AI—including AI safety research—these judgment calls are the main levers for people's own values to influence outcomes.
If people defer to AI without noticing, then "human in the loop" might provide less assurance than people expect.”
@MiaAI_lab ·
Opus 5.5 on Extra or higher feels like a different model.
Extremely strong and capable. It's hard to go back once you use it...
@AndrewCurran_ ·
Sam Altman's one-person company prediction is coming true. Anthropic just bought one. The founder of Mithril Medicine used Claude to take a muscular dystrophy antibody treatment from concept to proof-of-principle in mice in 18 months, with just himself, Claude, and part-time consultants. The technical report is great. I will quote:
'The week opened with the handover of two cell-based assays, cell adhesion and flow cytometry, within our team, while eleven questions remained open with the assay CRO. In two sentences, one of us asked Claude to review the latest emails between the outgoing assay lead and the assay CRO and to recommend next steps for the handover, and gave it the link to the handover folder. Claude returned a bottom line after 96s and a full plan with five drafted emails after 18 min. For the full plan it read six email threads, the handover folder, two slide decks and two reports, all in the mailbox and file store to which we had given it access. Four working days later, nine of the eleven questions had an answer in writing and the other two had been left to pilot experiments. The adhesion screen was thereby set out in writing. The CRO drew the plate map to the format that our team had requested, Claude compiled the run order and the material calculation from the production data, and one of us reviewed and approved them.
On day 4, the production CRO delivered its data package, a report of 341 pages, which one of us forwarded to Claude in two sentences, with the request to review it, share observations and recommend next steps. Claude returned a bottom line after 2 min 40s and, once it had read every page, a visual synthesis of all 113 constructs after 7 min 53s. The synthesis arranged the constructs on the grid of 14 by 8 arms. Claude marked 82 constructs for screening as is, seven for additional polishing, and the 24 constructs of three laminin-binding arms to be held back from the cell-based assays on production criteria. Each of the 113 constructs thus had a proposed next step in less than 8 min.
Claude had recommended binding kinetics for all 113 con-structs, and on day 7 one of us challenged this plan. Claude defended it within 27s, arguing that affinity would show whether the three arms merit further work. One of us overruled this recommendation and set the three arms aside on production criteria. Their material remains in storage. Four minutes later, Claude had revised its memo, the working page, the run order and three unsent drafts from 113 to 89 constructs. It had also applied the ruling to three dependent steps in the unsent drafts, each flagged for our veto. The exchange took 7 min 36s from challenge to new plan, and every affected email was still an unsent draft. After the ruling, eight emails drafted by Claude went out on the same day to three CROs and a consultant.
Over the whole week, Claude drafted 17 emails, each in answer to a request from one of us, and we sent 12 of them, each after editing. With its drafts, Claude listed 41 points, such as tube fills, shipping conditions, recipients and assay settings, for which it had found no written statement of ours, so that we could confirm, change or strike each before sending. We kept some, struck others and changed one. Among the passages that we kept were proposals for controls. One draft proposed a comparison of polished and unpolished material on one sensor chip, to size the error that impurities cause in the rest of the panel. Another requested a buffer control diluted like the most dilute sample, so that any effect of the buffer appears at its largest. These proposals concerned the measurement of constructs that had already been produced, and each reached a CRO only after one of us had reviewed and approved it.
@hammer_mt ·
Instant purchase
@MiaAI_lab ·
Out of all open models right now, I'd say DeepSeek V4.1 Flash 1017 is the best one yet, especially considering its relatively friendly size.
Now I'm waiting for the next GLM / GLM Flash.
@felixrieseberg ·
The coding interview is dead, at least for my team at Anthropic. I've seen enough to know that it no longer provides signal. I removed it from loops this month.
I have a spicy, optimistic take - I think there's an opportunity for junior engineers: https://felixrieseberg.com/hunger-beats-tenure-advice-for-junior-engineers/
@wesbos ·
Custom macro configurator using webHID
@levelsio ·
And...it's crawling out of its bottom! 😍
@DeryaTR_ ·
I fully support these efforts from Anthropic or any AI lab. We need to do whatever it takes to accelerate treatments. There's still a mountain or two to climb, but we must advance superintelligence & treat all diseases within a decade. Even a year's delay means millions dying!
@Voxyz_ai ·
Found the fix!
Codex writes code 𝘄𝗮𝘆 𝘁𝗼𝗼 𝗱𝗲𝗳𝗲𝗻𝘀𝗶𝘃𝗲𝗹𝘆. Even a tiny change gets null checks at every layer.
Spell it out for Codex: check data once where it comes in, then just use it. Write one-off logic right where it's used instead of splitting it into a function ahead of time. Skip try/catch wherever you can.
The team behind a 𝟮𝟮𝟬𝗞-𝘀𝘁𝗮𝗿 open-source coding agent on GitHub wrote these rules into their own AGENTS.md. Copy them straight into yours 👇
- Validate unknown values once at the boundary that owns them. Pass typed values inward instead of repeating `typeof value === "object"` and property-existence checks. - Do not defensively revalidate values already guaranteed by a schema, constructor, or internal type. - Do not extract single-use helpers preemptively. Inline the logic at the call site unless the helper is reused, hides a genuinely complex boundary, or has a clear independent name that improves the caller. - Avoid `try`/`catch` where possible
@deanwball ·
Amazing!
@theo ·
Every time I install an app update on my phone and it DOESN'T add iPhone Duo support, I feel a deep disappointment
@Miles_Brundage ·
I’m a centrist in that I thought Obama’s Cheesecake Factory joke was funny and also I like Cheesecake Factory
@petergostev ·
Dating in SF is crazy
@NVIDIAAIInfra ·
Quantum computers have a noise problem, and some of the world’s most powerful supercomputers are helping us tackle it.
Our researchers used 8,192 NVIDIA Grace Hopper Superchips at @fzj_jsc to test a leading fault-tolerance approach at record scale, so designers can refine quantum hardware before it’s built.
Learn more: https://nvda.ws/4euDxHm
@dani_avila7 ·
Ok this is really cool
@natolambert ·
People underestimate how much money is left to get thrown at sovereign ai
@trq212 ·
feels like there needs to be a new way to distribute opensource software such that people can modify via agents but also get upstream automatically merged in
coding agents can do this but they won’t do it automatically
@tszzl ·
this is one of the most positive developments in the history of machine intelligence