I’m 44. 3x Founder with 6 kids building an AI pharmaceutical company in Texas. Looking to hire AI-pilled psychos in jobs that aren't normally AI-pilled. Attorney. Lobbyist. Grant writer. Growth. Ops. PR. CS. EA. Reply with: I’m [age] I do [job] How AI changed [thing you do]

We were approached by a PE-backed healthcare RCM company whose senior billing analysts were answering the same denials questions by hand every week, from spreadsheets pulled off the ERP. Six weeks later, their technology leader was testing a working AI agent hosted securely on their own cloud servers. The agent answered 59 out of 60 test questions perfectly, matched their most important financial calculation to the penny, and gave the analysts back 15 hours a week. No private patient data ever left their secure system. The build ran in these seven steps, in this order: 1. Define what is correct before giving the AI instructions. We collected 60 real questions the analysts receive, paired each with the correct answer from the company's reports, and checked them by hand. Every change from that point on was measured against this fixed scorecard. 2. Lock down the data with read-only access to the database and a strict list of tables the AI is allowed to look at. This ensures the AI cannot accidentally change or write over official records. 3. Remove personal identities before the AI sees anything. Patient details were hidden at the database level rather than just asking the AI to ignore them in the prompt, which is the best way to ensure strict healthcare privacy compliance. 4. Force every answer into a strict format. If an answer doesn't fit the required layout, the system blocks it instead of sending an unpredictable, unstructured response to an executive. 5. Automatically grade every test run. We set up a system that checks the AI's output against our original scorecard, so a software update on Tuesday can't silently break the answers on Friday. 6. Keep the AI model easy to replace. We set up the system so the client can easily swap to a different AI provider if needed, and we built a dashboard to track costs, which run about $200 to $300 a month at this size. 7. Let the client do the final testing. The technology leader ran the tests himself, asking the hard questions he thought would break the system, and the AI held strong at 59 out of 60 correct. Plugging in the AI model was actually the fastest part of the project. Building the safety, security, and testing rules around it is what took the full six weeks. That is the Velocity Framework applied to a daily business process instead of just software development.

Building a product went from months to a weekend. Getting paid for it did not. Building is fast, marketing and growth are hard. Today we fix that. We are excited to announce CREEM 2.0 and that we have raised 7M EUR in total funding @ProductHunt launch and details below 👇

I'm done staying quiet about Higgsfield. Time to expose the truth.

AI video creation is getting more powerful, but only a few have figured it out. Thats why I built Videoclaw. I want it to be so easy, you don't need to figure it out. Just prompt, and watch it edit, generate and create video :) Today we launch. The video was made entirely with Videoclaw, blending human and generated footage. To show how truly easy it is, prompts and proofs for each project are in the thread. Try it for free. Oh, and there’s one more thing 👇

After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev • 20-200x faster • 40-400x cheaper (w/ output tokens free) • Frontier composable intelligence optimized for decisions AFAICT the shortest path to AI-based economic revolution

Today, I'm excited to announce a huge leap forward for search Now Shopify stores have *more powerful search than Amazon* By adding AI Autocomplete to their search, users can filter by typing, which increases sales significantly by helping people find what they want faster. Just by adding it from the Shopify App store. Search has been the same the last 15 years... until now. This new proprietary technology upgrades your store's search, by combining the intelligence of LLMs with the speed of search, all grounded in your store's catalog. Search used to be a 5 step process: 1) Type 1-2 words 2) Scroll 50 results 3) Filter down to 5 results 4) Find what you want 5) Buy Today, we've cut out 2 steps, making search far faster: 1) Type exactly what you want 2) Find what you want 3) Buy And while this is the best search, this is just the start... soon, people will see that this real-time intent layer by @magicX_ai is the foundation of: more powerful assistant experiences, generative interfaces, and eventually handsfree interfaces. If you know any Shopify stores, please tag or forward to them. They can add it in 1 click in the Shopify App Store: apps.shopify.com/ai-autocomp…

Today we’re unveiling Odyssey-3, a big step forward for foundation world models. It can control robots, power humanoids, drive cars (on the roads of India!), train AIs, pilot drones, and even play video games. We can’t wait to see what intelligent systems it enables.

Meet Moonphase Flip: a low-distraction AI screen on the back of your phone that keeps track of your to-dos and takes actions before you ask. Threads below 🧵 Explore more: hellomoonphase.com/s/17e982