Yuriy Dybskiy

Was looking for a designer to make a new website for @puma_ai and decided to try latest vibecoding options so played around with @antigravity and Codex from @OpenAI. Guess which one is which:

@elwatto i guess i need to go to Japan to pick them up too

@html5cat currently novablast 5. thinking of getting megablasts and superblast 3 in a couple of weeks in japan

one year of consistent running, starting from full sedentary dad. First time ever hitting ‘high’ vo2 max. It works.

@hthieblot 👨🏻‍💻✨

Only incredible founders can reply to this tweet

@visualsofearth1 @RyanResatka gorgeous! Where is this?

California’s wildflower superbloom

@hagaetc @KyleSamani I think the question is which particular app is best to try it in

@KyleSamani You can track the data here dune.com/hashed_officia…

Supposedly 100M x402 transactions for $30M in payments volume Where can I try this myself?

@levelsio @tbpn @photomatt @X you’re absolutely right 🫡

I think @tbpn should report on this with their image post :D BREAKING: WordPress founder @photomatt solves @X's AI reply bot problem

@sama contextmaxxx

I would like a single word for this phrase: "throw it into the maw with every bit of context I can think of".

@wesbos 🧊

Only cool people can reply to this

@OpenAIDevs @OpenAI ty ty 👨🏻‍💻✨

@html5cat @OpenAI great ship

finally rebuilt puma.tech to be a static pages vs Notion site to improve performance and drastically reduce number of request/js thanks @OpenAI Codex

i'm at level 6/7 and it's super fun arcprize.org/tasks/ls20

@mikeknoop is ARC-AGI-4 going to be StarCraft 2 and RDR2?

ARC-AGI-3 and ARC Prize 2026 are now live with $2,000,000 in prizes! As of today, version 3 is the world's only unsaturated agentic intelligence benchmark. Humans score 100% and frontier AI scores ~0%. Play here: arcprize.org/arc-agi/3 While no single version of ARC is definitionally AGI, our aim with the ARC-AGI Series is to continually produce useful scientific benchmarks which identify large remaining gaps between Humans and Frontier AI. At some point, we'll be unable to, and then we'll have AGI. Our new benchmark consists of over 100 novel game environments encompassing nearly 1,000 levels. Notably, test takers are given no explicit goals (other than to win) and must explore the environments to acquire goals, understand rules, develop strategy, and ultimately execute a plan to win. ARC-AGI-3 is a test of agentic intelligence. Beating this benchmark requires on-the-fly world modeling and continual learning to adapt to evolving environments. To score 100% AI must beat all of the games as efficiently as the human baseline (e.g., the number of actions taken to win). An ARC first, this gives us a formal comparison of AI reasoning efficiency vs humans. Version 3 carries classic ARC design principles: core knowledge priors only, private test sets to measure generalization, and it's fun! Every benchmark we release is an experiment and I believe this new version will provide strong signal towards increasingly autonomous AI agents. Prior versions of ARC held strong predictive power for important AI moments. Version 1 only saw progress with the release of AI reasoning models in late 2024 and Version 2 only began seeing progress with the advent of agentic coding models in late 2025. Version 3 is expected to signal when AI agents can become economically useful in more open-ended domains (beyond highly measurable domains like coding and math). There are a few other important design changes for ARC-AGI-3. The public set is now a "demonstration" set, not a training set. And unlike prior versions, the private set is now explicitly designed to be Out Of Distribution (non-IDD) from the public demo set. This is to mitigate targeting and because LLMs can now generalize over IDD splits using AI reasoning. Frontier models have made great progress over the past year. So much that several industry leaders have suggested we may already have AGI. Part of the ARC Prize Foundation mission is to provide accurate public sense finding and we strive to reduce false-positive claims. To this end, we've updated our testing policy. Going forward we will only verify scores outside of the official Kaggle competitions from AI systems with high commercial usage or are 100% open source. We're also adopting a stateless client scoring philosophy to ensure humans and AI are tested under identical conditions. The goal of these changes is to reduce the amount of developer-aware targeting (whether incidental or intentional) and provide clear signal if actual AGI progress has occurred. The Foundation also has a goal to inspire AI innovation which is most likely to come from the community. We've seen dozens of startups using ARC as a tool for showcasing their ideas - a few have fundraised serious capital based on their ARC results. To support this we're launching a new Community leaderboard. While scores for this leaderboard can't be Verified, and you should explicitly not trust these scores as an accurate measure of AGI progress, we will curate the best ideas and promote them. This year I expect we will see rapid progress on the ARC-AGI-3 Community leaderboard and the best ideas will eventually migrate into frontier models and onto the Verified leaderboard. Finally, we’ve partnered again with Kaggle to run two competition tracks for ARC-AGI-2 and ARC-AGI-3. This will be the last year for Version 2. When we launched the first ARC Prize back in 2024, I committed to running the Grand Prize until it was beaten. So for the ARC-AGI-2 track we will be paying out the Grand Prize to the best team, no matter what, in order to honor this commitment. In accordance with the Foundation mission, to win any prize money you must open-source a reproducible solution. We raised the standard for open source to include training. I'm excited to produce a truly open solution as a final send off for the ARC-AGI-1 and 2 format. Focus is now on ARC-AGI-3 (we've even started work on Versions 4 and 5). As always, I'm honored to have the opportunity to steward attention towards AGI progress. I'm also super grateful to the incredible ARC Prize team - including our core engineers, game designers, and human testers - led by @GregKamradt without whom we would not have this incredibly useful benchmark. See you on the leaderboard!

@harris # of series As you helped raise don't think Bridgewater would allow graph with non 0 starting point for Y axis tho

Here's a fun game we used to play at Bridgewater. Guess the chart. Winner gets...satisfaction.

drop the “Data” - we’re moving to Kyoto 🏯

@NotionHQ @ishverduzco i can haz jacket plz 🐾

@ishverduzco 👀

The vibes at @NotionHQ 🤌🏽

impressive lemonade stand you got there, David 🍋

@feross @zeeg @jfrog @SocketSecurity Wat. Such insecure take from jfrog

When the CEO of a $600M+ ARR public company calls out your startup directly, your team and customers deserve a response. @JFrog's CEO published a post today calling @SocketSecurity a "fragile, commercialized illusion of security" that "wraps open source scanners." This isn't the kind of discourse that makes our industry better. But since he named us, here are the facts. The attack he references -- the Trivy/Aqua supply chain compromise -- is one Socket helped expose. Our threat research team independently identified the OpenVSX extension attack on March 2, the 75+ compromised GitHub Actions tags on March 19, and the poisoned Docker Hub images on March 22. He's citing our work to make his case against us. On the core question -- who's actually finding supply chain threats -- the public record is clear. JFrog's research page lists ~5,000 findings across their entire 18-year history as a company. Socket discovers ~10,000 malicious packages *per week*. We've identified ~250,000 unique supply chain attacks. These numbers are all public. We publish our research, our detections, and our threat data publicly. Anyone can evaluate the work. We report our findings to the registries, where they end up protecting JFrog's own customers through OpenSSF. Scanners find known CVEs. Socket finds attackers. Those are different problems, and conflating them is either a mistake or a choice. JFrog's SEC filings show security is 7% of their FY2025 revenue. 93% of their customers aren't buying their storytelling either. Back to building.

“who are the best ppl in SF to talk about AGI?”

SF not messing around