Today, we share a breakthrough on the planar unit distance problem, a famous open question first posed by Paul Erdős in 1946. For nearly 80 years, mathematicians believed the best possible solutions looked roughly like square grids. An OpenAI model has now disproved that belief, discovering an entirely new family of constructions that performs better. This marks the first time AI has autonomously solved a prominent open problem central to a field of mathematics.

Slackbot has single handedly transformed how I work. Truly exciting to see more people get their hands on it

Marc @Benioff says he started Salesforce after he got a vision while swimming with a pod of 100 dolphins in Hawaii: "I was one with the pod." "I pay attention to my dreams. I pay attention to my visions, my insights. "Salesforce started because I was swimming with a pod of dolphins outside the coast of Hawaii, where I live. About 100 dolphins. I was one with the pod. And when I was one with the pod, all of a sudden in my mind, I saw this vision of what Salesforce could look like." "I went to Larry Ellison and said, 'Hey I think I'm going to quit my job.' He gave me $2 million, and we started Salesforce. That was 26 years ago."

Sharing an interesting recent conversation on AI's impact on the economy. AI has been compared to various historical precedents: electricity, industrial revolution, etc., I think the strongest analogy is that of AI as a new computing paradigm (Software 2.0) because both are fundamentally about the automation of digital information processing. If you were to forecast the impact of computing on the job market in ~1980s, the most predictive feature of a task/job you'd look at is to what extent the algorithm of it is fixed, i.e. are you just mechanically transforming information according to rote, easy to specify rules (e.g. typing, bookkeeping, human calculators, etc.)? Back then, this was the class of programs that the computing capability of that era allowed us to write (by hand, manually). With AI now, we are able to write new programs that we could never hope to write by hand before. We do it by specifying objectives (e.g. classification accuracy, reward functions), and we search the program space via gradient descent to find neural networks that work well against that objective. This is my Software 2.0 blog post from a while ago. In this new programming paradigm then, the new most predictive feature to look at is verifiability. If a task/job is verifiable, then it is optimizable directly or via reinforcement learning, and a neural net can be trained to work extremely well. It's about to what extent an AI can "practice" something. The environment has to be resettable (you can start a new attempt), efficient (a lot attempts can be made), and rewardable (there is some automated process to reward any specific attempt that was made). The more a task/job is verifiable, the more amenable it is to automation in the new programming paradigm. If it is not verifiable, it has to fall out from neural net magic of generalization fingers crossed, or via weaker means like imitation. This is what's driving the "jagged" frontier of progress in LLMs. Tasks that are verifiable progress rapidly, including possibly beyond the ability of top experts (e.g. math, code, amount of time spent watching videos, anything that looks like puzzles with correct answers), while many others lag by comparison (creative, strategic, tasks that combine real-world knowledge, state, context and common sense). Software 1.0 easily automates what you can specify. Software 2.0 easily automates what you can verify.

Salesforce has signed a definitive agreement to acquire @SpindleAI, a leading agentic analytics platform. The acquisition brings aboard deep expertise in multi-agent systems and high-performance data applications to accelerate Agentforce: sforce.co/4hLOpRA

the best movie on context engineering

Data Science teams are often stretched thin, and spend more time managing dashboards rather than discovering verifiable insights. Spindle AI changes this!

Sorry you don't like our codec implementations in VLC. Looking forward to seeing yours.

"don't write or run any code, just explain possible approaches and their tradeoffs"

Pre-training vs. Post-training — the new Nature vs. Nurture

Please tell us about the time you most successfully hacked some (non-computer) system to your advantage Soham: You're not going to believe this...

All of that is about money, not about the quest of building great things. You will often feel a malaise eventually when that is the focus.

???

ostensibly strange architecture decisions can usually be explained by: 1) decision maker has not been on-call much in their life 2) decision maker has been on-call a lot in their life

EXCLUSIVE: Star Health is a $1.4B revenue insurance company whose CISO sold ~31 million Indians' data from salary to PAN card to a Chinese hacker for $43k. Ever wondered how these things happen? Here's a breakdown of the events "allegedly" with video proof. 1. Amarjeet Khanuja, CISO of Star Health, reaches out to xenZen through a referral from denol on encrypted chat app Tox on July 6, 2024. 2. xenZen says yes and they negotiate and land on $28k for customer data on Monero (crypto). CISO doesn't know Bitcoin is a bad idea for this nor has ever used escrow. 3. CISO sends hacker login credential and an API endpoint w/details on their proton mail. Hacker sends money and gets it. 4. On July 20, CISO says I can also give you all the claims data. They agree on $15k and repeat the above. 5. Five days later, hacker says his access was revoked and CISO says "You've taken 5TB and I want $150k now because senior management wants a cut." 6. Hacker asks for a refund and gives final warning. 7. xenZen posted a sale listing on BreachForums for diplomatic passports of India (unrelated). 8. Sep 25 is when the starhealthleak website drops with 2 Telegram bots for customer and claims data. From private sources, xenZen says he's bought and sold data from Indian companies before. Attached is video evidence which is unlikely to be spoofed. The media did not care until I posted about it yesterday. Star Health responded by legal threats against Telegram and Cloudflare and a forensic investigation. People in power in India (and perhaps elsewhere) will sell your data in a heartbeat. Why? No one seems to care.

Just confirmed: @sama named O1 after the @O1Visa

im an alien of exceptional ability

OpenAI: pip install openai and set OPENAI_API_KEY Anthropic: yea same but s/openai/anthropic/g Google: oh boy. ok so you have a GCP account? no? ok go set that up. and a payment method. now make a "project". SURVEY POPUP! k now gcloud auth. wait you have the gcloud CLI right–

I can see why gaming from age 15-24 takes away time from other pursuits…but really till age 13-14, video games are my most treasured childhood memories. I wouldn’t want to take that away from a child. I’d rather trust their ability to discern responsibility and avoidance

Nothing has improved my life more than my wife’s best friend marrying a loser