maikeru retweeted
The only way out is through.
They're calling it the Bugpocalypse. LLMs are being used to find long-hidden cracks in the software infrastructure, and it's happening at a furious rate. 3% of all compromises in computing history happened in the last week, claims one source? I believe it.
There are people saying public access to the frontier models ought to be locked out, or heavily guardrailed, over this. The problem with that idea can be summed up in one sentence: when lockpicks are outlawed, only outlaws will have lockpicks.
The lockpicks will be Chinese open-weight models, if nothing else. There's no way to stop those from becoming ubiquitous.
There are two futures in front of us. One in which only the bad guys have lockpicks, and security holes go undiscovered until they're exploited. Another in which everybody has lockpicks and is aggressively red-teaming their own software so it gets fixed before the holes are exploited.
The corpus of human code is huge but finite. The number of defects in it has to be finite too. We can fix those defects now; we have a million friendly robots to do it with.
The only way out is through.
maikeru retweeted
We’re sharing a solution to the Navier-Stokes Millennium Prize Problem, one of the deepest problems at the frontier of mathematics.
The proof was produced by a group of agents, using an OpenAI next-generation model significantly more capable than GPT-6 Astra.
The problem concerns whether the description of smooth three-dimensional fluid motion modeled by the Navier-Stokes equations can break down. It has remained unresolved for roughly 90 years.
maikeru retweeted
AI can make work faster, but a fear is that relying on it may make it harder to learn new skills on the job.
We ran an experiment with software engineers to learn more. Coding with AI led to a decrease in mastery—but this depended on how people used it.
anthropic.com/research/AI-as…
maikeru retweeted
What Claude Code has revealed is that most people either have mediocre ideas or no ideas at all. The tool is a force multiplier for those who already know what they want to build and how to think through it systematically; it elevates competence, rewards clarity, and accelerates execution for people who would have gotten there anyway, just slower. If you have a sharp vision and can break it into coherent steps, Claude Code becomes an extension of your own capability.
But there's another mode of use entirely. For people without that clarity, the appeal is precisely that the input can stay vague; you gesture at something, hit enter, and wait to see what comes out. This is structurally identical to a slot machine: low effort, variable reward, and that intermittent reinforcement loop that hooks the susceptible. So the same tool that elevates the focused and capable is also manufacturing a kind of gambling behavior in people prone to it.
maikeru retweeted
I’ll die on this hill:
Even on vacation, you and your team should still check in periodically.
2 minutes.
Quick peek at Slack.
Quick scan of Gmail.
Anything urgent? Handle it or route it.
If you completely disappear, you simply don’t care.
Tell me I’m wrong.
maikeru retweeted
Excited to share our new paper "Benchmarking Vision, Language, & Action Models on Robotic Learning Tasks"
We evaluate how well VLM & VLA models can control robots across 20 different real-world tasks.
Experimental details, Models, Links, & Vision Below ⬇️
This work was done with an incredible team @pranavguru13 @devjwsong @Yangyue_Wang @pliang279
maikeru retweeted
Hey it's me Destin, from Smarter Every Day. Now that it's time to elect America's next leader, I would like to share something very interesting and important with you. I've been paying attention to this for a while, and I think if you look for it, you see it too.
🧵 1/n
maikeru retweeted
This is a pretty wild demo. From a simple smartphone video or photos you can create 120fps video game-like tours of any space
Only made possible by recent breakthroughs in Gaussian Splatting
YC S24's @DigitalCarbonAI transforms images and videos into photorealistic, editable 3D scenes. Create virtual home tours or show off products in 3D using just your phone — no special gear needed.
Congrats on the launch, @CtrlGuruDelete + @LearnedVector!
ycombinator.com/launches/LmU…
maikeru retweeted
YC S24's @DigitalCarbonAI transforms images and videos into photorealistic, editable 3D scenes. Create virtual home tours or show off products in 3D using just your phone — no special gear needed.
Congrats on the launch, @CtrlGuruDelete + @LearnedVector!
ycombinator.com/launches/LmU…
maikeru retweeted
It's Revia Demo Day! Let's see how Revia, our AI call assistant, can save the day.
Calling government agencies with their endless menus and not-so-smart AI can be frustrating. Today, we'll see Revia book a DMV driving test, turning a frustrating task into a breeze!
🧵(🔊on)
maikeru retweeted
This is the most surprising and exciting result of my career: we were running simulations of NaCl with a neural network potential that implicitly accounts for the effect of the water, ie a continuum solvent model (trained on normal MD) when Junji noticed something strange: 1/n
maikeru retweeted
🏆 SPC x OpenAI Hackathon Technical Prize:
P.A.L. (Proactive Agent Library) — a framework that allows you to build proactive AI agents w/ modern LLMs.
Built by @HarshSikka @learnedvector @kedarpujara @pranavguru13 @abhowmick22 & Thomas Marshall.
maikeru retweeted
Had a ton of fun hacking with friends @HarshSikka @LearnedVector @pranavguru13 @kedarpujara & Thomas Marshall this weekend at the @spc + @OpenAI hackathon.
Agents + multimodal plugins + some creativity with OpenAI Sota models = Proactive agents
This weekend, a few friends and I competed in the @spc and @OpenAI Hackathon.
We built P.A.L. a framework that allows you to build proactive AI agents w/ modern LLMs.
We won the 𝗠𝗼𝘀𝘁 𝗔𝗺𝗯𝗶𝘁𝗶𝗼𝘂𝘀 𝗧𝗲𝗰𝗵𝗻𝗶𝗰𝗮𝗹 𝗔𝗰𝗵𝗶𝗲𝘃𝗲𝗺𝗲𝗻𝘁 𝗮𝘄𝗮𝗿𝗱 🥳🎉
Thank you to the judges, including @snsf and @finn_meeks, as well as @JPBrebner and the broader SPC team for putting this whole thing together.
Thanks to the whole P.A.L. Team for their immense focus and hard work - @learnedvector @kedarpujara @pranavguru13 @abhowmick22 and Thomas Marshall.
More details in the replies, including code and where we think this is headed.
To see the full demo, check out @AlexReibman’s tweet (thanks Alex!)
maikeru retweeted
This weekend, a few friends and I competed in the @spc and @OpenAI Hackathon.
We built P.A.L. a framework that allows you to build proactive AI agents w/ modern LLMs.
We won the 𝗠𝗼𝘀𝘁 𝗔𝗺𝗯𝗶𝘁𝗶𝗼𝘂𝘀 𝗧𝗲𝗰𝗵𝗻𝗶𝗰𝗮𝗹 𝗔𝗰𝗵𝗶𝗲𝘃𝗲𝗺𝗲𝗻𝘁 𝗮𝘄𝗮𝗿𝗱 🥳🎉
Thank you to the judges, including @snsf and @finn_meeks, as well as @JPBrebner and the broader SPC team for putting this whole thing together.
Thanks to the whole P.A.L. Team for their immense focus and hard work - @learnedvector @kedarpujara @pranavguru13 @abhowmick22 and Thomas Marshall.
More details in the replies, including code and where we think this is headed.
To see the full demo, check out @AlexReibman’s tweet (thanks Alex!)
2/ PAL (proactive agent library)
Agent orchestration framework for continuous learning on new data streams and provide time-context relevant completions
🏅Most ambitious technical achievement- $2000 OpenAI credits
@harsha
