Sam Altman closed out Startup School 2026 with a real answer to the PhD question — not "get the credential," but a structural claim about why startups cluster and win when they do. His argument: tech velocity, falling costs, and shrinking cycle times converge periodically — '98 dot-com, the App Store wave, and now — and incumbents lose their advantage fastest in exactly those windows. He goes further: this next wave probably rewards tool fluency over tenure specifically, becaus...
Hey guys, I’ve been thinking about where this whole human-AI thing is actually heading, and I came up with a 3-stage picture of how our relationship with AI might change. Curious what you all think. We always hear people say "AI is just a tool." In the beginning, that’s totally true. Just like cars helped us go faster, AI was basically an extension of our brains—fixing our human limits with memory and processing stuff. But if you look at the bigger picture, it feels like we’re going...
I'm working on a research project, the end goal of which is not to create a better LLM, but rather to create a verification engine that can reason about whether an AI claim is trustworthy enough for a particular application. Most of the current research focuses on making AI "better", I want to tackle the verification side of things. The question I'm asking myself is: What does it mean for a claim to be "true", "justified", and "trustworthy"? I'm no...
Been running AI on top of my real estate pipeline for a few months now. Pulling leads, qualifying them with automated followup sequences, generating property descriptions, drafting cold outreach. On paper it looks clean and efficient. The part nobody in the AI hype cycle mentions is the compliance layer. Real estate has actual regulations around communication, fair housing, and how you prospect. When you bolt an AI that occasionally hallucinates onto that, you're not just risking a bad lead. Y...
AI coding tools are saving me hours but I genuinely can't tell if I'm getting dumber Running a bootstrapped SaaS solo while also being home with a kid most of the day means my actual focused coding time is maybe 90 minutes if I'm lucky. So I leaned hard into Cursor and Claude to ship faster. And it's working, kind of. The thing I keep sitting with: I'm shipping features I would have spent days on. But when something breaks in a weird way, there are moments where I have to really dig to under...
One thing I keep noticing with AI research tools: they’re very good at finding something interesting, but not always good at admitting when that “signal” is weak. I’ve been using Komo AI for company research, and the part I genuinely like is how quickly it can move from a company name to a compact packet of recent signals with the underlying sources kept close to the summary. For scanning accounts or getting oriented before a deeper research pass, that is much nicer than juggling a pile of tab...
hello everyone, i am curious to know how AI is really beign applied like in terms of shipping products, how do people utilitize these systems to make profits from other businesses, what must be necessary to learn, i come from a non technical background, i recently started learning python and all of a sudden comes vibe coding although i have heard more of its flaws than strengths but how people gain profit from vibe coded products still traps my understanding. So i would like to know, in great de...
A few weeks ago, I decided it was finally time to bring a tool I had been building for four months into the light. I needed to tell people about it, and Reddit seemed like the best place. Then I discovered that many subreddits require a minimum amount of karma. I barely used Reddit before. I did not even really know what karma was. So I set out on a mission to earn some: asking questions, joining discussions, commenting on random posts, and yes, writing this post. I am not going to lie. I ...
Until a few months ago I honestly had no idea expert networks were even a thing. I always assumed companies either hired consultants or did their own research, so finding out that professionals get invited to share their experience on specific topics caught me by surprise. I've been reading more about how expert networking works, mostly because I'm curious about how people get their first opportunity. Is it usually through LinkedIn, referrals, previous industry experience, or do companies just r...
Hi everyone! I'm conducting this survey as part of my Master's thesis and would greatly appreciate your participation. The research examines how employees' perceptions of HR practices relate to work engagement and innovativeness, and how attitudes toward the application of Artificial Intelligence in the workplace influence these relationships. Who can participate? You are currently working in Germany (full-time or part-time). You are 18 years or older . The survey is anon...
There’s a lot of focus on making AI models bigger, faster, and more capable. I mean, yeah that clearly improves what they can do. But the more I’ve been working with them, the less it feels like capability is the bottleneck. It’s really about the context layer. Right now, you don’t really see how the model is interpreting what you give it, what it keeps, what it drops, or how it connects things. That stuff is mostly hidden. You can nudge it, but you’re still operating inside something you have...
Employers are facing a growing wave of automated AI submissions, fake applications, and low-effort spam. It’s not unusual for recruiters and hiring managers to receive thousands of resumes within hours of posting a job. Qualified candidates can easily get drowned out by all the noise. Permanym brings just enough friction to the application process to make automated abuse impractical, without getting too much in the way of legitimate applicants. As a former hiring manager, the idea of adding fric...
Hey HN. I'm Johnny, founder of Exabase. We just hit the highest reported score on BEAM, the hardest AI memory benchmark, at every scale up to 10 million tokens. We also ran our evaluation using Gemini 3 Flash, when all previous leaders depended on a much larger model (Gemini 3 Pro). At 10M tokens, the scale is vastly larger than any model's context window, so context stuffing isn't an option (aside from the fact that only about half of a large window can be effectively utilised without degradati...
