SquishyCoder — d/acc

Well. This is disappointing. I just canceled my premium subscription because @X rugpulled me and took away the hexagonal profile and NFT verification. It was literally the reason I started paying to begin with. #web3 #Crypto x.com/SquishyCoder/s…

I didn't drink at my birthday party so I could start crushing PR's all day today. I have 99 PR's to work through from the community, but a hangover aint one First big feature of the day: a chief security officer to audit your attack surface (informational only, please hire a pentester for real production code and mission critical situations!)

@JTLonsdale Pretty sure he’s going to get the nod next year for another five years, but after that will be wild. Anything could happen in the competition for succession. Hoping America, Japan and allies can be stabilizing enough that internal PRC chaos doesn’t turn into external aggression.

@SquishyCoder I was starting to spend time to find ways to ally between US and China - many were - but then they let a true commie get in charge and pushed things back for themselves, and our dynamic, by decades. Terrible self-own. Now CCP is just our enemy. Hope one day it’ll be different.

This is true, they also cracked down on scores of their top entrepreneurs, eliminating many and terrifying others not to build too successfully. No smart Chinese billionaire wants a 2nd or 3rd huge co. Something much of the left wants to do in the USA, but fortunately cannot.

@technxgga @Omniasciree @Govindtwtt Yep, very important IMO, especially for backend and systems dev.

@SquishyCoder @Omniasciree @Govindtwtt Code reviews are a thing

Software engineers invented AI for one sole purpose: To do front end development.

@dmartkc @ctdonath @Chris_arnade Now find one year during your entire lifetime when the rate of criminal incidents per million trips in Singapore was as bad as any of those numbers. Then try Taipei. Then try Seoul. Sorry, bro. Chicago and even DC are cooked and will be until getting serious about violent crime.

@ctdonath @Chris_arnade Why use statistics (3.7 criminal incidents per every 1 million trips in the DC area) when I could instead argue with the semantics of a take or use anecdotal reports to justify sweeping conclusions about an entire issue?

Every transit nerd — the urbanist who tell us over and over how much they love public transportation— has to realize until you stop this sort of crap, and make stopping it your number one goal, you and all your train love are doing nothing but annoying everyone else.

@dmartkc @Chris_arnade Live in east Asia for a while and you’ll realize it’s a completely solved problem.

I agree we should stop incidents like this, but actually doing it at scale seems like someone trying to say they’re going to stop all road rage incidents. This certainly isn’t the daily experience of most transit riders even though videos like this go viral. DC has noticeably increased security presence on trains, but short of security checkpoints at every station how does a large system actually prevent all anti-social people from ever using it? Seems likely to require more societal level solutions like additional mental health resources and crisis responders, etc (which I’d guess most transit advocates support!)

@Tapeseee @elonmusk He could become president of Mars!

@elonmusk I might say Elon has a promising future as a presidential candidate 💁‍♂️ Just saying

I would like to offer to pay the salaries of TSA personnel during this funding impasse that is negatively affecting the lives of so many Americans at airports throughout the country

@DevCreatorWorld @TechLayoffLover Nothing this account posts is true. It’s all doom porn that brings down huge numbers of people so he can earn a small amount of money doing writing he doesn’t even really enjoy. It’s pure cancer.

@TechLayoffLover Even if parts of this are true, I don't see how recordings alone replace experience.

A senior architect at a major cloud provider just discovered something that made his blood run cold For 8 months, IT has been screen recording every coding session under "productivity analytics initiative" 3,247 hours of his work. Every keystroke. Every API design decision. Every debugging session. Every architectural trade-off conversation. His manager kept praising him for "documenting best practices" and "knowledge sharing sessions" - he thought he was being helpful Turns out those 47 "knowledge transfer" sessions weren't for the team They were feeding everything into training datasets His prompt engineering workflows, his system design patterns, his troubleshooting methodologies - all catalogued and indexed Two L4 engineers in Bangalore now have access to his complete decision tree library They're shipping features 60% faster than his old team of 8 senior engineers The kicker: last month's performance review praised him for "successfully training the offshore team" His manager literally said "you've made yourself scalable" Yesterday he found the Confluence page titled "Knowledge Extraction Playbook - Phase 2" His name is in the "Completed Extractions" section The offshore team lead just got promoted to Principal Engineer His badge expires Friday But hey, at least he helped "democratize institutional knowledge" The architect who built the system is now redundant to the system

@deedydas @RahulCapitalLLC How about this one with swarms of open source models, like Minimax M2.5? github.com/pass-agent/loo…

@RahulCapitalLLC ngmi

I’m not exaggerating, I hear from so many big software cos which don’t use Claude Code/Codex. CTOs are asleep at the wheel. Engineers are typing code by hand. Fixing a bug a day. Like it’s 2024. If youre at these cos, demand change or leave. Now. You’re in for a rude awakening.

@deedydas @cclark How could you know that if you “vibe coded” it? Did the prompt responses actually bring up those kinds of implementation details?

@cclark Novel, for some definition of the word. It didn't import anything, the search and board eval etc is in Rust

I just "vibecoded" a Chess master (~2250 ELO) from scratch that runs locally on a Mac in Rust. I used to play chess semi-competitively, and I'm flabbergasted that you can just speak a 98% percentile chess engine into existence.

@Omniasciree @Govindtwtt The feedback loop is much faster with frontend, which makes LLMs a bigger win. Since you can instantly see when they generate a UI bug, the bug is cheap. If they create a bug in your data pipeline, you might not notice until your system falls over under a heavy workload.

@Govindtwtt Frontend?😂 I think u meant backend

@andrewchen That’s not true. Email came BEFORE the web. Someone as old as you should have heard of BBS and there were schools and companies, with email, too. AOL’s famous “You’ve got mail!” clip was recorded in 1989. CompuServe first offered email in 1979.

Web 1.0 came with new channels: - email, search, link sharing, etc Web 2.0 too: - feeds, creators, viral invites, etc Mobile: - app stores, SMS invites, vertical vid, mobile ads What about AI? I’ve been complaining that AI hasn’t come with much. But we’re seeing a big growth channel opening now: Products that are built as APIs/CLIs that can be pulled into new projects by Codex/Claude on the fly Maybe the “AI-native hotel app” doesn’t mean a mobile booking app with an AI chat panel. It means a CLI that can book a hotel for you, that an AI agent can pull into a bespoke answer or project or into code. Bolting on an AI chat panel is this generation’s weak form of AI. Maybe the full reinvention involves making it agent-first not human-first and once you start looking at it that way, a lot of existing products suddenly feel mis-specified. they’re built as destinations, but agents don’t want destinations. they want capabilities. composable, callable, reliable capabilities. So instead of “go to Expedia” or “open the app,” the future interaction is more like: an agent assembles a workflow on the fly. it pulls a flight search tool, a hotel booking tool, maybe a weather model, maybe even your personal preference graph. none of these are full products in the traditional sense. they’re more like endpoints with taste and state. This flips distribution completely. historically you win by owning the surface area. seo, app store ranking, homepage traffic. in an agent world, you win by being the default callable primitive. the thing that shows up again and again in agent-generated plans because it works, has clean interfaces, and returns structured outputs. distribution shifts from “top of funnel” to “top of call stack.” And the crazy part is this might actually compress product surface area dramatically. the best products might look more like tight, extremely well-designed CLIs with opinionated defaults rather than sprawling UIs. almost like the stripe api moment, but for everything. imagine if every vertical had a “stripe-level” primitive that agents preferentially use. there’s also a weird inversion of brand here. humans used to choose brands. now agents will. so the brand becomes partially machine-legible. reliability, latency, error rates, schema clarity. you can almost imagine “agent seo” where the ranking factors are things like success rate across thousands of agent runs, or how easy your tool is to integrate in a chain-of-thought execution loop. This also suggests a new kind of moat. not just data or network effects, but integration depth with agent ecosystems. if claude or codex or openclaw learns that your tool is the safest way to accomplish X, it gets baked into prompts, templates, maybe even fine-tunes. you become a default. and defaults, historically, are insanely sticky. The contrarian take is that most current “AI features” are a local maximum. chat panels, copilots, assistants. they’re transitional. the real end state might look closer to invisible infrastructure that agents orchestrate. the ui is just a debug layer for humans to peek into what the agents are doing. so maybe the new growth channels for ai look like: - being callable - being composable - being reliable at scale in agent loops - being embedded in agent templates and workflows - being the default primitive in a given domain and if that’s right, then the question for any new product isn’t “what’s the ui” or even “what’s the killer feature.” it’s “what’s the minimal, highest-leverage capability we can expose such that agents will repeatedly choose us when building something new.”

@CodveAi @andrewchen I thought Instacart, Doordash, Airbnb, etc, all spent a fortune on ads

@andrewchen this is theYC thesis in a nutshell. the best companies don't advertise - they spread through organic usage. when your product sells itself, you've built something defensible.

lots of AI cos starting to experiment with paid marketing so here’s my take: Paid acquisition is a tax on your product's defensibility. the moment you can't out-spend the incumbents and competitors, you die. build channels that get cheaper as you grow or you're just renting your growth

@rohanpaul_ai Agents are more motivated to sell houses quickly than to maximize the sale price. Their monthly earnings is: (commission rate) * (sale price) * (number of sales in a month). Of course they prefer a higher sale price, but they won't do twice the work just to get it 20% higher.

Fortune: A Florida homeowner used AI to bypass agents and sell his property for $100K above estimates. He managed the entire pricing and staging strategy through simple text prompts. "The AI planned the most granular aspects of the homeselling process. It gave tips on how to update the property, even suggesting which walls to repaint. And it told Levine when to schedule home viewings to work around his schedule. The father of three ultimately showed his home to 15 prospective buyers, one-third of whom submitted an application." The seller only used a human lawyer for the final closing documents.

@einarvollset @patrick_oshag @WorkMJ It’s almost line they’re competing against each other.

@patrick_oshag @WorkMJ Mm. I wonder if this is more of a stable equilibrium thing. If it became easier and rewards stayed the same, more people would do it, increasing competition until more failed, reverting to the equilibrium? I dunno.

A crazy image

@paulg @tekacs In the other direction, Chinese, doesn’t add spaces between words. Ancient Latin didn’t, either.

@tekacs In effect, prose is a compressed form of bulleted list, and people who are sufficiently good at reading prefer the compressed form.

The fact that AIs tend to answer you in bulleted lists tells us something important, though somewhat depressing: people can't read. They don't do this by accident. What you're seeing is an implicit portrait of the median user.

@AuraLeeHarvey @micahsb @Noahpinion The keys are: 1) don’t tolerate violence or crime 2) don’t support homeless drug addicts 3) treat racial groups in an equal and meritocratic way 4) encourage excellence in schools 5) encourage civil debate between political opponents

@micahsb @Noahpinion We’ve all contributed. We need to see that, and be more cooperative with one another. That will still the thunder of leaders who try and bring out those excesses as a way to move their ball down the field.

People who didn't experience the pre-2014 liberal American culture don't have any idea how awesome it was. It was creative, fun, gentle, tolerant, funny, idealistic, accepting, supportive, and smart. It was why I didn't permanently move to Japan in the 2000s. Now it's gone.

@micahsb @justuseKsharps @Noahpinion Hmm… I wonder if MAGA would exist if not as a reaction to those with your extreme views and support of violence. I don’t think Trump could have won without the backlash against your tribe.

@justuseKsharps @Noahpinion More like a sport though really. Shhhh, I'm Maga hunting.

@micahsb @90sDemInAmber @Noahpinion I’m not partisan and agree with him, though. Americans have some extreme views and those he mentioned (plus encouraging homelessness, drug addiction and crime) are the ones you see in far left cities.

@90sDemInAmber @Noahpinion lol. Nice partisan bullshittery ya got there

@ee546636 @Sach_R81 @business It think he has very good intentions but is locked in a constant internal debate about what move is optimally "good".

@Sach_R81 @business He’s a strange guy, I don’t think he has good intentions

Peter Thiel’s Founders Fund is backing a company bringing AI to cow herding at a $2 billion valuation bloomberg.com/news/articles/…