Samantics

DOGE dropped the largest Medicaid dataset in history couple days ago. Everyone's posting screenshots of the top billers. Nobody's digging into WHY the numbers look like that. I downloaded all 227 million rows. Here's what I found underneath. 🧵

Oh no. Let me guess the DM: """ Hey Brett. We would love to offer you $5K for a post about our new upcoming launch of Supercomputer. All we ask for is a retweet with a quote 'The fact that this is one prompt and one agent is genuinely wild.' Thanks - Higgsfield """ I would immediately stop working with you if i was one of your clients just for this obvious grift

There has never been, and there will never be, a more exciting time to be a creative. The fact that this is one prompt and one agent is genuinely wild.

Respectfully what you said is word salad. You are saying "Don't connect ChatGPT to your bank accounts because you may not have proper security measures on your chatGPT account" How is that a viable suggestion you are making? You think the people vulnerable to this type of vector aren't just reusing passwords anyway across their systems? Bank accounts are connected to a bunch of different apps. Unless your position is connect nothing to nothing then you don't have a defensible position here

@inferencepoint Here’s a breakdown of how these attacks function. My goal is to make it harder for me to find a single point of failure.

Just worked w/ @CNN on a piece about why I wouldn't upload full financial docs (tax docs, statements, etc) to AI tools due to leakage & hacking risk. I don't recommend connecting bank accounts to AI tools. It becomes a 1 stop shop for attackers looking to drain your accounts🤖💸

@ChatGPTapp Rocket money shaking in their boots rn

A preview for Pro users: a new personal finance experience in ChatGPT. Pro users in the U.S. can securely connect financial accounts, see where their money is going, and ask questions based on the information they choose to connect. Your full financial picture, now in ChatGPT.

@thsottiaux Let's make `codex --worktree` happen please. Worktrees are so easy to manage on the app but CLI needs some love too to make it easy

Who said you can't? Just pay API pricing People get this wrong constantly. Anthropic just doesn't want to handle people complaints about rate limits, credit exhaustion, bad performance because of harnesses they don't control. They have made a bet that they can provide the best experience and cache control mechanisms on their harness. And if something breaks or regresses they have all the data they need to fix things. None of that is true when you just use another harness

Why does Claude not want us to use 3rd-party harnesses? Are they afraid we'll use fewer tokens than usual?

Credit to @xai for giving them access to compute. Probably has helped Anthropic capacity plan a lot better and ease up their micro management on what you can use your sub for

@_Evan_Boyle Lol not the dunk you think it is. Some no good no name project probably also shipped it. If you have dropped the ball on literally every other feature that people want you don't get to claim a win for a sub feature

Congratulations to everyone who has discovered /goal in the last two weeks. Copilot CLI shipped `autopilot` mode in January.

@MKBHD No but it should fill out 99% of the forms every payment site has and just let me enter credit card details or pay with apple pay

Ok genuine question: Would you actually trust an AI with your credit card to execute this in one click?

Theo couldn't crack it if he started YouTube today. Lol I wouldn't worry about it. He rode the covid hype train and then pivoted into talking about AI drama for the last 2 ish years off that audience You guys actually have interesting ideas and content. You just need to work on the packaging and distribution a bit. GPT 5.5 video was good. Far more useful for serious builders than anything Theo has put out in 2 years

blocked and reported for being a gigantic asshole who is also, unfortunately, correct about this

@DavidPattt AI slop posting has gone insane on this platform

Class of 2026 10 years later

@DanielMiessler @trq212 Take only makes sense pre AI. And only if you are choosing to write it yourself. The second you are asking AI to write you're deferring the cognitive overload to your reader and you should do that in a way that's easy to consume at least. HTML is the best medium there

This is really cool thinking from @trq212 here, but I think I disagree with the solution. He makes a great point about Markdown being more difficult to share and communicate ideas with, because formatting and visuals can make things super easy to understand. My problem with the approach is that, by trading editabilty for readability, we’re separating we humans even further from the creation process. I value Markdown because I value text. And I value text because I see it as one step away from thought. I believe thinking is the one thing we should be careful not to outsource, and I worry what this idea smuggles in is a major step toward making our creations opaque to humans. Not just AI's creations, but ours as well. The reason I value Paul Graham so much is because of the idea compression work that goes into writing super clean prose. It's difficult to write clearly because it requires thinking clearly. Text makes your ideas naked, and I like that. - What is the problem, exactly? - What should we do to solve it? - Why is our solution better than alternatives? I love the challenge of crystalizing this kind of critical stuff in pure text before any technology is involved. If we're not writing that text ourselves, and then editing it, it starts to feel a lot like bringing a strong robot to the gym. I worry that if we vibe-think to AI and have it spit out amazing HTML, we're instantly disconnected from the idea. Like where did the idea go? It started as vibes and got put through a woodchipper and turned into someone else's HTML. Can I see it in 4 simple bullets? Can I stare at it? Can I grapple with it. Can I tweak it? It's an idea. I need to be able to wrestle with it. Of course we can ask the AI to summarize its brilliant HTML document into four bullets, but we'll have lost through compression and expansion some percentage of the original. Maybe I'm being overly emotional here. I just feel like if you didn't put the hard thinking and writing work into the original idea, and then maintain it in a format that's easy for humans to read and edit, then you have somehow surrendered something Holy to the machines. I say this as a total AI maximalist. But I get the point he's making, and I think it's super valid. It's hard to explain or convince people of things with a giant text file. Formatting massively helps. Images massively help. Even an interface or a video or something. So we're synched on that. I just think it might be better to come at the output we both want in a different way. - MARKDOWN: Easy for humans to write, hard for humans to read. - HTML: Hard for humans to write, easy for humans to read. Maybe the solution isn't moving the first step to HTML where it becomes more opaque to both agents and humans (plus the versioning issues Thariq talked about). Maybe the solution is something crazy like document pairing: like you have the thought file and you have the presentation file(s). The proposal is to ask AI to just write HTML, right? Well why not just have a separate but linked file for that? One is for crystal-clear human creation and sync between human and AI. Simplicity, clarity, precision, and human editability. And then AI can produce whatever from that. Images, diagrams, videos, or whatever. And if you want, yes, a full HTML file that contains all of them. And that can be what you use to present or share the idea with audiences. (Plus there's the fact that some file formats are literally directories, which could be shared with lots of related content, and then there's also things like .mdx that allow for richer content in Markdown, etc.) I hate the idea of multiple files, but I think it's far preferable to losing the transparent, editable connection to the idea that you get with text. Plus, the better and cheaper AI gets, the more trivial it will be to have the core thought file plus n-number of associated versions or formats that are useful for different audiences. Basically I think it's much easier for AI to make a rich and shareable version of clean, editable thought, in the form of text, than it is for humans to stay connected with ideas as opaque HTML. And I think the human thought-to-text connection is the most important thing to preserve. Still thinking it through, however, and massive thanks to @trq212 for the push for all of us to evolve on this.

@badlogicgames For Claude models the split is crazy. 30% input. 70% output. Literally the flip for gpt models

TIL about "caveman mode" to "save tokens". how many tokens in a session are actually model output? i think i'll become a gardener.

@GergelyOrosz @clairevo Just look at Whoop or Skool communities. It's filled with "AI automation agencies" because someone told them building with n8n was the future

One thing I am certain will happen (probably is happening): A massive industry emerging selling snake oil. Aka people with 0 real experience posing as “AI-native consultants.” Happens in every gold rush. Pay attention to ppl who actually do the job like @clairevo instead tho

@badlogicgames This tracks. I just looked at my input vs output token split and GPT 5.5 is like 70:30 skewed towards reading vs Opus which is 70:30 the other way

they did not and it's all just vibes as usual. the variance is still annoying.

I have tried just about every model under the sun from all the major providers. Cognition's is the only one I haven't been able to so far. A new customer is surely what you're looking for? 😃 I run billions of tokens a month in my team and would be happy to bring some of that business over if it sticks

🎁 We're giving away 5 Windsurf plans ($250 credit each)! Try SWE-1.6 — @Cognition’s latest fast and intelligent agentic coding model, powered by Cerebras. In a side-by-side with Claude, the speed difference is clear. More iterations, faster fixes, better code. 💬Comment why you want access to enter. Five winners will be selected at random within 48 hours.

@RhysSullivan Yes I checked directly through a proxy

does changing between thinking levels on Claude / Codex bust the cache?

@mitsuhiko Bait. Now post some actually good PR's so the community can learn what to do!

A selection of great PRs that were submitted to Pi — a thread.

@GergelyOrosz Sticking it to Sam 😃

So let me get this right: 1. Anthropic bans xAI from using Claude (to stop them from perhaps distilling Claude for their own model) (...) 2. xAI gives up ~a quarter of its DC capacity for Anthropic to rent and run Claude A win for Anthropic no doubt. What's in it for xAI tho?

Bots on X love the name Sophie