michael froehlich

@marcklingen what language?

Duolingo is finally good now Lots of interactive voice elements that get you to speak quickly And the owl continues to be hilarious Want to learn a new language and thought about getting a private teacher / classes, but I think I’ll just see how far I get with Duolingo

@karrisaarinen beautiful emdashes

You know a company is starting to lose the plot when they start publishing nonsense comparison pages. Who is juggling Linear, Jira, Rovo for AI, and Loom? 😂 (we still haven’t published a single comparison page)

@OpenAI live demos >>> polished apple keynotes things might fail. still 100% more authentic

It all comes together in 10 minutes. openai.com/live

@chrija how cool is that! @chrija how does it taste?

New blog post: Creating a New Protein Crop From a Forgotten European Plant When we started talking to Padraic and Mike, I knew nothing about the seed business (despite having been a seed investor for almost two decades, LOL). Here's what convinced us to bet on the aardaker, a tuber combining the best of potatoes and soy! medium.com/point-nine-new…

@RhysSullivan dataleap.ai maybe?

alright i'm wanting to try background / proactive agents, looking to learn what peoples setups are like rough things on my mind of random agents i want: - i want an agent that watches a github repo like an sdk and tells me how it changes - i want an agent that whenever i get a slack message it decides whether to page me, start investigating, etc - i want an agent that every 4 hours figures out things i need to reply to / follow up on - i want an agent that watches every github commit for how those changes are doing in production - i want an agent that works as my growth agent running experiments, monitoring posthog, tweaking my landing page marketing copy etc - i want an agent that monitors all my production traffic for any anomalies these don't need to be super aggressive on responding to triggers, some form of cron is fine i've tried codex automations but something didn't really click with their implementation for me also somewhat tried hermes but same thing of just something didn't really click with the implementation eve and flue maybe fit into this category but just seem so high friction to set up for these use cases, my dream UX is just sending a prompt to an agent and getting back a proactive agent setup i do see this being one of the biggest unlocks for making life easier when working with ai but haven't found a setup i like yet

@MaxMynter could not imagine a better person for the CI pipeline

friendgroup just planned out our to-be-founded hippie commune. I‘ll be responsible for our surfboards and the CI pipeline

@KevinVanCott pretty cool!

I might have just figured out how to integrate Web Workers with TanStack Table V9... So... multi-threaded TanStack Table coming soon? This would theoretically unlock huge UI performance wins for specific large tables that use heavy client-side computations for filtering, sorting, aggregation, etc. In a normal setup, the UI can freeze for a second or two while a large filter or sort is computating, but as seen here, the UI stayed 100% interactive while sorting 1 million rows with a slower alphanumeric sorting function. I want to know if this would be useful to anyone.

@ben_issen what's weavy?

Made this whole campaign in Weavy. 2 weeks from idea to Times Square.

@mattpocockuk how do you ingest them? self-built?

Doing my first ever experiments with a personal, entirely agent-managed Karpathy-style wiki X, Discord, Gmail are all being ingested into it every few hours This is the knowledge base that will serve as the environment for all of my future loops

@milindS_ lmk if you need help setting up gbrain. took me a while! monodrive.ai

So PSA for anyone who wants to build this A *fantastic* starting point is Hermes + Gbrain You can message it on Slack It has ALL company knowledge Can be invoked in parallel Can use Hermes to code OR have it drive Claude Code or Codex You can connect everything to it But also Your data remains yours You can swap models, personality, and add custom features whenever you want I'm sure @Teknium and @garrytan would agree

@azeem and there goes my weekend

The GenAI economy has generated $110 billion in sales over the past 12 months. It is growing fast. On an annualized basis, the revenue run rate exceeds $175 billion. These numbers took us several months to construct, and as far as we know, it’s the first bottom-up, deduplicated measure of consumer and enterprise AI spending across the full stack. We are releasing this research today in our first The State of the AI Economy report. intelligence.exponentialview.co

@karpathy trying hard not to say “we copied openclaw”

This is a new paradigm for interacting with Claude that is significantly more "inline" with all the other human activity org-wide. Once you do all of the under the hood engineering work to make this "just work" (e.g. across tools, integrations, compute environments, memory, security, etc.), Claude basically joins the team in a seamless way - you can talk to it as you would talk to a person and it can help with a very large variety of workloads. Imo this is the 3rd major redesign of LLM UIUX. The first paradigm was that the LLM is a website you go to, the second was that it is an app you download to your computer. This third one is that it is a self-contained, persistent, asynchronous entity with org-wide tools and context, working alongside teams of humans. It really takes a while to wrap your head around it, but it works and it is awesome.

@JensMannanal make things agents pay for

@froehlichmmm And having payment authorisation

i think we underestimate how good ai models will be at making good decisions can ai run companies? probably prerequisite: having the right state of the world

@theo openai money printer goes brrrrr

If you're curious how I managed to do over $20,000 in inference on the last 48 hours, here's a video all about it. Spoiler: loops are really powerful

@tonychenxyz cool stuff. models seem to be getting good at judgment calls in noisy, partially observable, delayed-feedback system

Today, agents execute isolated tasks. Tomorrow, agents will steer complex decisions across long horizons. Introducing CEO-Bench, a first step to measure "Steering Intelligence." In CEO-Bench, agents are asked to run a simulated startup for 500 days. ceobench.com

@jhleath real sick!

we're thrilled to be working closely with the E2B here! give your sandboxes powerful, infinite disks that synchronize to your s3 bucket! an important primitive for folks who: are building agents and need to hand them context or running RL and need to compose results easily into an S3 bucket

@EvaSpannagl @AITinkerers @OpenAI @felixkrrr @aloglute @annabellschfr great demos!

full room at our first @AITinkerers x @OpenAI event in Berlin yesterday we saw everything from customer support tools, a Whatsapp answer bot answering tech questions of your dad… and even a robot playing games. the variety was incredible. you can just build things. thanks to our great co-hosts @vietdle, Langdock. More to come.

@midjourney sci-fi stuff! curious - how does the preparation work? Putting people in a water tank sounds like there is substantial on/offboarding time?

A technical dive inside our new "Midjourney Scanner"