Alec Wantoch

These were good times

everyone talking about agentic payments for web requests, why wouldn’t model providers start acting like token banks? trade token credits directly instead. if everyone is adopting LLMs, it might end up a valid form of payment. stablecoins make sense too but more overhead

So I have been testing the windows version of the codex app. No bluescreens so far which is nice. Heads up, it does require your modem to be dialed up at all times. That's a bummer because I was expecting the AI stuff to be included with my local MSDN install.

@christofsalis @steipete why does this go crazy hard tho

@Mike_Scully_ that’s where we’re at: beemflow.com

The biggest AI opportunity nobody is talking about? Mid-market companies. $5M to $50M revenue. Too big for cookie-cutter solutions. Too small for enterprise consulting. Drowning in manual workflows. Have the budget. Have zero AI expertise. They know they need to change. They have no idea where to start. That's AI paralysis. And it's everywhere right now. The market is wide open. All it takes is the right offer and a cold email sequence.

architecture still matters boys and the models can't do it yet without you same features, 5k LOC vs 93k. insane compression by finding the right architecture

@Grok440 @mert use claude in chrome for ui testing !

@mert I raw dog lol, I have not opened an IDE actually in I think 8 months. Claude even deploys... would love to have it somehow test a ui though... just kinda waiting or maybe someone with a pointer :)

kids these days seem to have such complex AI setups anyone else just raw dogging a local folder via terminal and claude/codex code?

@mert same vibe dawg x.com/alecw/status/2…

@jspeiser strategically it would make sense to have performance shares here as well, no? especially if equity wasn't pre-negotiated. something like "ok i understand you want deeper alignment with the upside of the company, here's what we can do based on these targets"

A founder in our community owns 100% of his company. His best salesperson just asked for equity. The person has been a major driver of growth and frees up the founder's headspace more than anyone else. His questions to the group: 1. Phantom shares vs real equity? 2. Should I tell him we're building toward an exit? This is one of the most common conversations founders have and one of the least prepared for. 3 mistakes I see: - Giving equity out of guilt instead of strategy - Not telling key employees about the exit timeline - Making the structure so complicated that the exit itself becomes harder The right answer is almost always: phantom equity with a clear vesting schedule, tied to a liquidity event. Keep the cap table clean. And yes tell them about the exit. Alignment is more valuable than surprise.

isn't it crazy this was 10 years ago

@richa_lq idk this seemed like a normal deep convo to me I thought it gave Jensen an awesome opportunity to frame the strategy, his answers were so on point and articulate 🔥

Dhwarkesh done played devil’s advocate too close to the sun. I genuinely donno how he handled it - it just kept getting progressively more uncomfortable as the video continues.

@theo pshh javascript? vibecoded web assembly obvs theo cmon

Why are you still using React when you can vibe code something better in a day?

@notiansans @runwayml seriously amazing !

What started off as a workflow test, just an idea, quickly had me locked into a flow state. 4 hours total from completely blank slate to finished video. I can’t believe we’re all so lucky to have this at our fingertips. 🔊 sound on 🔊 @runwayml

@kanjun great launch video !

Twitter’s algorithm is optimized for addiction, not for us. We deserve better. We’re releasing Bouncer today so you can take back control of your feed. Describe what you don't want, and Bouncer removes it. It’s free, doesn’t collect your data, and will be open source soon.

@theo this has been working great for me: x.com/alecw/status/2…

I want to control my agents from my phone. I'm sure you do too. What workflows do you want? Do you have one that works now? Tell me all about it.

everyone wants to automate the hard stuff there's no such thing as "claude make me a billion dollars make no mistakes" i just want to automate the boring easy/rote stuff that wastes time so i can get back to where my brain is actually useful

@sowmay_jain @laukiantonson this is cool but..handing your DNA to AI companies willingly? 😳

i got my whole genome sequenced two years ago and forgot about it. last week i told my ai agent (@laukiantonson) to dig up my DNA files • it dug up a two-year-old email • found the download link • pulled down 67 gigabytes of raw DNA. • rented a 32-core, 64GB machine for a few hours — total cost: $5 • aligned 21 million long reads to the human reference genome — 99.83% mapped • called 5.8 million genetic variants using a two-pass neural network • phased every variant — separated maternal vs paternal inheritance • annotated all 5.8M variants against ClinVar, PharmGKB, and gnomAD • corrected for population-specific bias in the medical literature • health risk map across 39 conditions flagged in every body system • drug compatibility guide for 141 medications color-coded by genome response • nutrient metabolism - 71 variants affecting absorption of vitamins, minerals, iron • traits, ancestry going back 40,000 years, neanderthal DNA breakdown $5 in compute. 8 hours. no bioinformatician. no doctor. just one instruction. we've genuinely reached a point where an ai agent can take your raw genome and hand you back a full personal health profile in a single shot. i had no idea this was even possible.

@datacology @samposthuma might want to check for 2.4GHz interference from some nearby devices

@samposthuma Does your trackpad not lag? Mine does and it’s annoying

it changed my life, work goes 2 times faster, it’s like having a laptop with huge screen where you can move the keyboard around without moving the screen, best 30$ ever spend

distribution has always been the moat of moats

@JoshKale part of me wonders whether this is just a model running without the safety layer, that they were capable of this already, but this boosts hype and enterprise value and influence at just the right time

This is big... Anthropic just announced a model so powerful they won't release it to the public out of fear over the damage it will cause 😨 Claude Mythos Preview found thousands of zero-day exploits in every major operating system and web browser... The numbers are hard to believe: > $50 to find a 27-year-old bug in OpenBSD, one of the most security-hardened operating systems ever built > Under $1,000 to find AND build a fully working remote code execution exploit on FreeBSD that grants unauthenticated root access from anywhere on the internet > Under $2,000 to chain together multiple Linux kernel vulnerabilities into a complete privilege escalation exploit For context: these are the kinds of findings that previously required elite security researchers working for weeks. Anthropic engineers with no formal security training asked Mythos to find exploits overnight. They woke up to working code the next morning. The results were so impressive Anthropic assembled Apple, Google, Microsoft, Amazon, NVIDIA, and seven other organizations into Project Glasswing: A $100M defensive coalition. They're not releasing this model publicly. Instead, they're racing to patch the world's infrastructure before models like this proliferate.

in a world of sycophancy, (healthy) skepticism is even more valuable