Dr Val

@julian_englert @tamarindbio @phylo_bio 🤯

We're officially releasing the Adaptyv API, which gives you and your AI agents access to our wet-lab to test your proteins experimentally! • Check out our demo and the docs here: agents.adaptyvbio.com • Check out how our partners @tamarindbio and @phylo_bio have integrated the Adaptyv API into their platforms We started Adaptyv with the idea that anyone should be able to test a designed protein, whether they have their own lab or not. Over the past three years we've tested tens of thousands of proteins from pharmas, AI for protein design companies, academic labs, alongside dozens of early-stage startups and individual researchers. Until now, all of that went through our Foundry portal or also email threads and Slack channels. We think the process of testing a designed protein should be as simple as calling an endpoint, so we built an API around the same infrastructure those teams already use, to make everything as accessible as possible. As AI agents will do more and more scientific work, it's important to give them the tools to access real-world experimental validation. To put it simply: AI can think but it cannot touch - we're giving AI access to the lab to validate experimental hypotheses.

@damianplayer @mcuban @tbpn 👇🏼

Mark Cuban (@mcuban) just laid out the BIGGEST opportunity of the next decade: helping companies actually implement AI. (bookmark/save for later) no-fluff breakdown below 🧵

@demirdjiantwins Banana

We replaced a $197K/year content team with Nano Banana + Linah AI + VEO 3 From concept to upload, the system now produces 300+ short-form UGC ads every month fully automated, faceless, and optimized for conversions. It started as an experiment: Could we scale ad creative without hiring creators, editors, or managers? So we built a machine that: - Generates 300+ AI UGC videos on demand - Auto-posts across 50+ TikTok burner accounts - Tests 500+ hooks, voices, and formats - Tracks winners & doubles down automatically The results? - 22M+ organic views - Dozens of videos crossing 200K+ each - Multiple TikTok channels hit FYP consistently - Shopify + affiliate funnels → $368,000+ tracked revenue Why it worked: - Zero creator management headaches - 100% organic growth (no ad spend) - Mass testing + rapid iteration - Linah AI + VEO 3 running nonstop We originally built it as our private growth engine. Now brands are asking for it. So I’m opening up the full playbook: The video automation system Distribution & scaling strategies The AI agents powering it Best practices for VEO + Linah AI Want your own Linah AI + VEO 3 UGC engine? RT + Comment “BANANA” and I’ll send the blueprint for free. (must be following)

@damianplayer Algo

BREAKING: X just exposed the EXACT algorithm that decides who goes viral… And I just reverse-engineered it into a custom trained GPT: → The exact formula for viral posts → FYP selection patterns (from X) → Engagement thresholds that trigger boost → Which signals get you shadowbanned Like, rt + reply "ALGO" and I'll DM it to you (must be following)

@codi_fy DuckDuckGo

@appealstoheaven @iruletheworldmo Yep

Grok has answered everything I’ve thrown at it satisfactorily. And it’s pretty complex work stuff that isn’t well documented. GPT has never. You hyped grok 4 and I think it’s not what you acted like it would be. So I have no reason to believe gpt5 will be so great. Gemini 2.5 Pro and Grok 4 are the top dogs right now.

just seen some early benchmarks for gpt 5 and it absolutely destroys grok 4. which i’m hearing was trained on all of the benchmarks.

@DerekNee Couldn’t find any detail re data policy - is the users data private or will be used for training?

introducing agent neo, with your complete knowledge. infinite steps. infinite memories. unlimited output. a SOTA agent that executes complex tasks in the cloud. superhumanly intelligent. upload thousands of your documents to give it your memories. your digital twin, with your entire cognition. no subscription. no limits. all yours. free. enjoy.

@prof_tariq @Ankesh290 @JpKayobotsi @flowith VHAGELIE

@DrVal0risation @Ankesh290 @JpKayobotsi @flowith Already used ☹️

can't keep up with the code requests so we're changing things up - everyone gets 7 codes to share! and, when someone uses your code, you get +2 days free. win-win! drop yours in the replies and let's get this community going

@Ankesh290 @JpKayobotsi @flowith Used it But here is another one: 419KZ2OH

@JpKayobotsi @flowith R154667I

@xykhdyjc12392 @flowith … all used (within a minute!)

@flowith EK1GOGTM 56WLB03D NO1L9MYM XOCLJNDI 6DLOGEQ2

@DerekNee Looking forward to trying it out 🤩

introducing flowith NEO. the SOTA agent i built for the future. infinite steps. infinite memory. endless output. cloud-running for a year straight. benchmark-beating intelligence. zero cost. zero limits. enjoy.

@NiohBerg 👇🏼

You can literally see her small brain break after he asked her that question.

@em @LiorOnAI everything is in flux and ai labs are trying to innovate as fast as possible

@DrVal0risation @LiorOnAI Interesting, otherwise they get caught up in some sort of "incest artificial content loop" that could harm the training set because its too artificial?

OpenAI quietly released their GPT-4.1 Prompting Guide. It's a must read if you're using agents or LLMs.

@em @LiorOnAI Kinda, (just my hypothesis) in a way they want ai to “learn” with “bad” instructions over time. So it can become more accessible for everyone. If you keep giving “perfect” prompts - with an intermediary step - ai does not get smarter even though you would get better outcome

@em @LiorOnAI … to gather more (messy) Training data - helps making the next model “smarter”

@LiorOnAI i don't understand why we can not just prompt the way we prompt and the LLM reads those prompting guides, corrects our prompts, and prompts itself in the "better way", can someone explain that to me?

@lisyarus Now do the universe

Ah, all the artifacts were fixed by replacing one '<' with '<=', duh. It all works now, particle life completely on WebGPU! There are 2^18 (~260k) particles here; my PC struggles after about half a million particles. Will probably try to optimize it a bit later.

@realEstateTrent The transaction cost of doing these type of things will approach zero 🤷🏻‍♂️

What happens when you then send your as-built drawings to the AI, and it puts the building plans together? What happens when the City has an AI tool that reviews those plans and provides instant comments? That crazy farfetched future is no longer farfetched. AI is here!

The leap I witnessed in AI this week has me shaken. If you've followed me, you know I don't really make statements like this. I've been hearing how AI will change everything for years, but there's been little beyond promises. Nothing that can be used now. That just changed:

@visegrad24 Is this a new show on Netflix?

BREAKING: Jackson Hinkle is speaking in front of hundreds of thousands of Houthis. He’s saying that “America and the Zionist entity fear Yemen more than anything else (…) and are waging war against God” He’s a traitor of America and the West

@JimFergusonUK 😳

🔥 INTERVIEWER: “Can’t we just plant more trees to fix climate change?” 💬 Bill Gates: “That’s complete nonsense… Are we the science people, or are we the idiots?” Well Bill… 🌳 Trees do absorb CO₂. 🌎 Nature does regulate climate. 💰 But your “solution” always seems to involve endless control, digital IDs, lab meat, and experimental vaccines. So we’ll ask the question back: Are YOU the science guy… or the other one? 👀

If you feel like shit, you’ll eat garbage! Plenty of research links poor mental health to unhealthy eating. Similarly, studies on self-esteem show that without self-value, people neglect their diet. “Preference” isn’t just taste; it’s shaped by emotional and psychological struggles often tied to systemic issues.

Why do low-income households eat so much garbage? It's not prices, nor the quality of available food. Leveraging supermarket entry, we can see a clear answer: It's mostly because they prefer eating garbage.