The average company buys Snowflake, then Fivetran, then a BI tool, then spends 6 months connecting them. We spin it all up in 30 seconds.

this is cool, but 95% of jev demos ignore the existence of agents. if you needed to fill this form in, saying "find my resume and fill this in" to an agent is enough to get the job done without a specialized tool

this is exactly what happened with Mintlify for us. I had 25% of my Claude max leftover on Friday night, I got an email from them saying they were increasing prices, so I replaced it in ~12 hours with about 25 minutes of actual effort.

This is cool, but I think too late. People are already used to getting *exactly* what they want from an agent. e.g. in @definiteapp we often see people take pictures of a sketch on a piece of paper or a screenshot from a few other tools (Stripe + Excel). Our agent has templates to start from, but ultimately writes a react app to give the user what they want. It'd be hard to get that experience in a framework like this.

trying to reconcile AGI timelines with this video games timeline

Looks like @bot just pushed a partial reset? I was near 100% weekly usage and now at 47%. Not complaining! Just wondering what happened.

GrokBot is pretty good. I got access via my Cursor account for $1,152/year, so $96 per month. You get a 8cpu, 16gb machine baked in, which would cost over $100 per month on AWS. Not to mention ~3.4 billion or so tokens per month.

I'm not going to care about open models until some blend of the below becomes true: 1. the labs stop offering max plans 2. really smart open models can easily be run on my mac 3. TPS (token per second) AND intelligence are gpt5.6 level on #1, it's nearly impossible for me to run out of codex tokens right now (I have 4 resets banked) and Fable 5 seems to be sticking around for the forseable future. I have virtually unlimited token usage for $400 a month, so open models being cheaper doesn't matter to me. on 2 and 3, benchmarks are showing some of the open models at around opus4.8 levels, which is incredible! But running them locally at anywhere near the TPS of cloud inference is far off. I can run a smaller (dumber) open model locally and get good TPS, but see #1, whats the point?

this is really one hell of a pitch: ✅ I'm too drunk to drive + homeless ✅ I *might* exceed very low expectations ✅ You should expect the law to be involved

taking June off

when Anthropic released Claude Design: "figma is dead" but Fable is so fucking good, they're killing their own products now. I was eyeing Claude Managed Agents when it launched, but we already built something better with Fable in a few days.

Fable 5 is Anthropic's most "honest" model to date. It tells you right in the response. I index all my local Claude Code sessions in DuckDB. I have 202,381 messages in the last 30 days alone. There's been a steady increase since Opus 4.6 in the model saying "honest". It probably shouldn't, but this bugs me. Should I assume most of the time you're lying and you're being honest in this one message? I was pumped in the first few hours of Fable where this had seemingly been "fixed". 100+ messages and no "honest" to be seen. But it was too good to be true. Within a few hours, Fable proved itself to be the most honest model to date. Analysis done in Codex because I need to save my Fable usage for real work 😅

do not try to reattach anything yourself

Every era is blind to something staring it right in the face