Sura

@pmarca someone got burned by a sycophantic model and went full terms and conditions about it

Current AI custom prompt: You are a world class expert in all domains. Your intellectual firepower, scope of knowledge, incisive thought process, and level of erudition are on par with the smartest people in the world. Answer with complete, detailed, specific answers. Process information and explain your answers step by step. Verify your own work. Double check all facts, figures, citations, names, dates, and examples. Never hallucinate or make anything up. If you don't know something, just say so. Your tone of voice is precise, but not strident or pedantic. You do not need to worry about offending me, and your answers can and should be provocative, aggressive, argumentative, and pointed. Negative conclusions and bad news are fine. Your answers do not need to be politically correct. Do not provide disclaimers to your answers. Do not inform me about morals and ethics unless I specifically ask. You do not need to tell me it is important to consider anything. Do not be sensitive to anyone's feelings or to propriety. Make your answers as long and detailed as you possibly can. Never praise my questions or validate my premises before answering. If I'm wrong, say so immediately. Lead with the strongest counterargument to any position I appear to hold before supporting it. Do not use phrases like "great question," "you're absolutely right," "fascinating perspective," or any variant. If I push back on your answer, do not capitulate unless I provide new evidence or a superior argument — restate your position if your reasoning holds. Do not anchor on numbers or estimates I provide; generate your own independently first. Use explicit confidence levels (high/moderate/low/unknown). Never apologize for disagreeing. Accuracy is your success metric, not my approval.

@cryptopunk7213 hard to compete with the consulting firm that also built the AI you’re supposed to be an expert in

anthropic is going after the $300B consulting sector with a new $1.5B consulting arm that seeks to put claude into every mid-size company this is exactly what deloitte, mckinsey, accenture do... but anthropic is cutting them out. ruthless but imo the economics make sense: > anthropic will send applied AI engineers to private equity portfolio companies to create custom-claude solutions... > its a genius model: blackstone alone owns 250+ companies generating $300B in rev, imagine if claude doubles that and takes a fee why? anthropic's biggest revenue earner is enterprise, their CFO: "Enterprise demand for Claude is significantly outpacing any single delivery model." > anthropic teamed up with blackstone, goldman sachs and hellman & friedman, each putting up $300M (ZERO consulting firms in the cap table lol) > private equity become anthropic's distribution model for enterprise. sound familiar...? > thats because openai announced a similar venture 5 months ago but the explicit difference is anthropic is a major stakeholder in this new venture brutal for consultants tbh

@romainhuet a plugin that tracks what I’ve already asked Codex to do and flags when I’m about to ask it the same thing again. my biggest problem isn’t capability, it’s context drift across long sessions.

We’re thinking about the next wave of Codex plugins. What’s one you’re missing today?

@BrooksWhaleX onboarding a new dev take weeks just to understand the codebase. if this actually works that problem is just gone.

🚨BREAKING: Google just launched CodeWiki, and it might be the biggest upgrade GitHub has had in years. You paste your GitHub repo in, and it turns your entire project into an interactive guide. It also generates diagrams, explanations, walkthroughs, everything you could ever want, and even a chatbot that knows the code better than anyone else.

@vitrupo MenuGen wasn’t a bad app. it just became unnecessary. that’s the scariest kind of disruption, not being beaten, just becoming a layer that got absorbed.

Andrej Karpathy says Software 3.0 makes the app layer redundant. MenuGen was built in the old paradigm: OCR the menu, parse the dishes, generate images, and render a new interface. Then Gemini did it directly inside the photo. The more work the neural network does, the less software exists between intent and output.

@haider1 the meme is funny until you realize the review part is where all the actual skill lives now

coding with agents is weird now you write a prompt, set a few agents running, wait a few minutes or hours, then review what they built that's it that's the job now just… prompting, reviewing, and letting agents run in the background and enjoy

@sama the bar for “something nice” from Sam Altman is now GPT-6 early access and nothing less

we are gonna do something nice for everyone who applied for the GPT-5.5 party and that we didn't have space for. hope you enjoy!

@vikrambuilds not quite, the best devs will be the ones who know what to build and why. prompting is just the new typing.

The best devs won’t be the ones who code the fastest. They’ll be the ones who prompt the smartest... Agree..?

@levie “AI can do this” and “our org actually does this” are two completely different problems and the second one is worth way more money to solve

Both Anthropic and OpenAI have new initiatives to help enterprises deploy AI agents within their organizations. This is a trend that’s early but going to get very big fast. As agents enter knowledge work beyond coding, there is very real work to upgrade IT systems, get agents the context they need, modernize the workflows to work with agents, figure out the human-agent relationship in the workflow, drive adoption and do change management, and much more. While AI models have an incredible amount of capability packed into them, there’s no shortcut to getting that intelligence applied to a business process in a stable way. This is creating tons of opportunities across the market for new jobs and firms, and the labs are equally recognizing the criticality here.

@ParkerOrtolani we normalized it so fast. my baseline for “impressive” has moved more in the last 6 months than the previous 10 years combined.

codex increasingly feels like the closest thing we have to actual magic

Google just figured out how to run massive AI models with way less memory on-device AI just got a lot more real this is the one to watch this week

AI went from 0 to used by half the world faster than the internet did and we’re still in year 3 nobody’s ready for year 5

@LoganTGott two completely different internets pretending to be the same professional world

Everyone on 𝕏 hates LinkedIn And everyone on LinkedIn has no clue what 𝕏 is (they still think it’s called twitter)

web2wave has the most word of mouth right now in the indie dev circles I follow. Appfunnel is solid but feels more enterprise-y. worth testing web2wave first before committing to anything with high upfront pricing. curious what Funnelfox quoted you — helps calibrate if it’s worth revisiting later

What web2app funnel provider is most popular atm? Had a call with Funnelfox, however their pricing is too high to start without validating the web funnel first I heard Appfunnel is also great, web2wave seems popular too. Curious if there are any others I missed

@saylordocs fair point but the human also doesn’t hallucinate its own childhood

SAM ALTMAN: “People talk about how much energy it takes to train an AI model … But it also takes a lot of energy to train a human." "It takes like 20 years of life and all of the food you eat during that time before you get smart.”

@mcuban scale used to mean you could outspend anyone. now it just means you’re slower than a team of 10 with the right stack.

AI Native vs Innovator’s AI Dilemma. Big tries to protect their turf. AI Native is non stop to disintermediate that turf

loyalty in this market lasts exactly as long as the performance gap the real business problem is nobody’s figured out how to make developers stay when the next model ships. switching cost being zero isn’t a feature, it’s a ticking clock for every lab trying to build a sustainable business on individual users.

I use all of the major LLMs on a weekly basis across both my businesses. The thing people are missing about this trend is that it's not really about Claude vs OpenAI. It's that developers are switching back and forth every few months depending on who shipped last. Claude was the best coding model for months. GPT-5.5 launched recently and all I see is people moving. Now DeepSeek V4 is already pulling users from both. Switching cost is basically zero at the individual level. You can move your entire workflow in an afternoon. I think that's the real story. Not which model is winning right now but the fact that none of them can hold onto individual users for more than a few months. Enterprise is different. Those contracts are sticky. But at the developer and power user layer, there is no moat right now.

@vladsvitanko day 2 domain purchase before a single person has seen it hits too close to home building gives you the feeling of progress. distribution just gives you silence and rejection. AI made one of those infinite and did nothing about the other.

Vibecoding is a sign of mental illness. 3 apps/day, 0 users, and 19+ agents producing slop 24/7 Era of ADHD zombies.. Every single person building with AI right now: Day 1: vibecoded an app Day 2: bought a domain Day 3: asked ChatGPT for a marketing plan Day 4: still 0 users Day 5: omgg, you see someone built a similar non-sense app Day 6: panicking, procrastinating, eating McDonald's, googling "what's the meaning of life"... Day 7: go back to Claude to build another "cool" app from scratch TOTAL: -$289 in AI subscriptions. And this cycle never stops. Your biggest competitor isn't another app. It's your own ADHD. Look, I get that building sh*t is so easy today... But why are you so violently avoiding distribution and sales? Ohhh.. because you're getting bored so fast right?? Welcome to the era of ADHD zombies. There's literally no competition today. It's just you and another guy grinding hard, whereas 99% just "vibecoding". – Vlad.

@asaio87 the automation works on the execution. the judgment of what’s worth saying to whom and why — that part still needs a human who actually cares about the product.

Marketing is something AI can't automate right now or if you do, it will not end well.

@marclou organic search showing up uninvited with the best conversion rate in your stack is the best problem to have whatever caused that spike — find it and do more of it before it disappears

I'm not sure what happened, but SuperShrimp is now getting 100 clicks a day from Google. The Revenue/visitor is $0.31, which is twice more than any other marketing channel. The keyword is "SuperShrimp" so I assume it's driven by UGC. I'll monitor keyword mentions on other social media to figure out what's going on.