Explore

Just got AI to make this table so people can easily get their countries flag for use in their bio: Country Flag Philippines 🇵🇭 Brazil 🇧🇷 South Africa 🇿🇦 Nigeria 🇳🇬 Kenya 🇰🇪 Colombia 🇨🇴 Mexico 🇲🇽 Argentina 🇦🇷 Saudi Arabia 🇸🇦 United Arab Emirates 🇦🇪 Indonesia 🇮🇩 Malaysia 🇲🇾 Thailand 🇹🇭 Vietnam 🇻🇳 India 🇮🇳 Türkiye 🇹🇷 Egypt 🇪🇬 Morocco 🇲🇦 Ghana 🇬🇭 Chile 🇨🇱 Peru 🇵🇪 Romania 🇷🇴 Portugal 🇵🇹 Spain 🇪🇸 Italy 🇮🇹 France 🇫🇷 United Kingdom 🇬🇧 United States 🇺🇸 Canada 🇨🇦 Australia 🇦🇺 New Zealand 🇳🇿 Germany 🇩🇪 Netherlands 🇳🇱 Sweden 🇸🇪 Norway 🇳🇴 Poland 🇵🇱 Bangladesh 🇧🇩 Pakistan 🇵🇰 Japan 🇯🇵 South Korea 🇰🇷

i'm doing a (free) media format conversion program (just a hobby, wont be big and professional like GNU) for macOS. It's not total vibe coding but I didn't realize how awesome AI could make stuff before it was better (in my opinion) then anything i've ever tried before. now i'm being semi-disciplined with it. Converts video, pictures, audio to/from any format ffmpeg backend (currently it has the fat proprietary codec included ffmpeg build which isnt open source compatible but literally supports everything) really slick looking interface that adheres to apple design philosophy has cool ✅, ❌and 🔄 simple coherent and useful interface that doesnt have all the nonsense and overly complicated options programs like this typically have. no cumbersom file dialog boxes, drag and drop and in-place output I just made the github repo public and haven't got around to including release builds or a trendy looking README.md yet https://github.com/jsarasin/MacConvert

The thesis falls apart because it confuses technical impressiveness, production cost, historical difficulty, and customer value as though they are the same thing. The fact that an AI can do in ten minutes what once took a skilled person three months is not evidence that the user is receiving "three months of value." That is literally what technological progress does: it collapses the cost of previously expensive capabilities. A spreadsheet is not worth the salary of an accounting department because it automates work accountants once did manually. A camera is not worth the labor hours it would take a painter to reproduce every photograph. A compiler is not priced according to how long it would take someone to hand-write the corresponding machine code. The entire point of technology is that past difficulty stops determining present value . And the "these people don't understand what they're getting" argument is backwards. Customers are not required to understand the provider's infrastructure in order to evaluate a product. Nobody buying cloud storage is expected to contemplate data-center construction, drive failure rates, power contracts, networking, cooling, redundancy, and engineering salaries before deciding whether $10 a month feels expensive. They look at the product being offered, the price being charged, and whether it satisfies their needs. If an AI company says, in effect, "pay us $200 per month and use our amazing model," then users will evaluate the service against that proposition. If ordinary usage under that proposition is economically ruinous for the company, that's a pricing and product-design problem. It is not evidence of a moral defect in the subscriber. The "$800 worth of tokens" argument is particularly confused. Even assuming that number is accurate, API retail price is not synonymous with the provider's marginal cost. A company charging one customer $800 for API usage does not mean it incurs an $800 expense when a subscription customer generates the same number of tokens. Retail pricing incorporates margins, capacity management, differentiated products, demand, capital recovery, and strategy. Treating retail API price as though somebody just burned eight hundred actual dollars in a furnace is nonsense. And even actual production cost doesn't determine consumer value. Suppose an inference costs $50 to run and produces something I value at $3. The fact that it cost $50 does not magically make it worth $50 to me. Conversely, an inference might cost fifty cents and save a corporation $100,000. Cost and value are different variables. The attack on "vibe coders" isn't much stronger. Yes, people can now produce software without understanding everything underneath it. So what? Programmers themselves have spent seventy years building abstractions specifically so that future programmers wouldn't need to understand everything underneath them. Most web developers don't understand transistor fabrication. Most application programmers don't understand CPU microarchitecture. Most Python programmers couldn't implement CPython. Most people using a database couldn't write a database engine. Most developers happily use enormous libraries containing code they did not write and could not reproduce. Every generation of abstraction lets people operate competently at a higher level while being ignorant of lower levels. AI is a more dramatic version of that process, but "this person couldn't have made this without the abstraction" isn't an indictment. That's what the abstraction is for. Of course inexperienced people will produce bad software with it. Inexperienced people produced bad PHP, bad JavaScript, bad Visual Basic, bad WordPress sites, bad Excel macros, and bad shell scripts long before LLMs arrived. AI lowers the barrier to producing software, which inevitably lowers the average skill of the population producing software. That tells you almost nothing about whether lowering the barrier is good or bad overall. The claim that AI output is typically "shallow, incomplete, and highly derivative" also sits awkwardly beside the complaint that users don't appreciate how extraordinarily valuable the models are. Which is it? Are these systems generating trivial slop that sophisticated people can easily see through, or are they delivering the equivalent of years of expensive expertise for pennies? The post wants AI to be unimpressive when insulting its users and miraculous when defending its price. You can't have both arguments whenever convenient. Then there's the bizarre idea that people should reserve frontier models for sufficiently "worthy" tasks. Why? The scarce resource belongs to the company selling access to it, not to some informal priesthood of enlightened users. If the provider wants to discourage expensive usage, it has all the normal mechanisms available: quotas, metering, tiered pricing, slower priority, API billing, or different plans. A user choosing the strongest available model for email summarization may even be perfectly rational. Perhaps it follows instructions better. Perhaps it makes fewer errors. Perhaps the user doesn't want to maintain a mental benchmark of twelve model variants and perform microeconomic optimization every time they ask a question. The cognitive overhead of constantly deciding which model is "sufficient" also has a cost. The post effectively says: you've been given this incredibly convenient general-purpose intelligence, but you're using it wrong unless you continually think about the compute economics behind every request. That rather defeats the purpose of making the technology convenient. The entitlement argument is similarly misplaced. If someone believes that paying $200 should give them unlimited usage when the plan explicitly promises limited usage, they're simply mistaken. But if the company markets the service vaguely as extraordinarily capable and broadly available, continually changes limits, or creates expectations that aren't met, customers are entitled to complain. That's how markets work. Customers don't owe suppliers reverence because supplying the product is difficult. The most revealing flaw, though, is the claim that AI has "ruined people's minds" because they no longer appreciate the effort that used to be required. By that standard, nearly every successful technology has ruined people's minds. People don't appreciate what it once took to typeset a book because word processors exist. They don't appreciate the labor involved in long-distance calculation because calculators exist. They don't appreciate navigation because GPS exists. They don't appreciate photographic chemistry because phones exist. They don't appreciate manual telephone exchanges, hand-drafted engineering diagrams, rooms full of human computers, or physically searching library indexes. Good. Civilization advances partly by allowing ordinary people to stop thinking about solved problems. There are legitimate concerns hiding underneath the rant. AI can create overconfidence. People can become unable to evaluate output they didn't understand well enough to produce themselves. Excessive dependence can cause skill degradation. Cheap generation can flood the world with mediocre material. Those are interesting problems. But none of them establish the post's larger thesis. What's actually happening is much simpler: AI has commoditized capabilities that were previously scarce, and people are beginning to treat those capabilities like commodities. That feels insulting if your intuition about their value was formed when they required rare expertise, enormous labor, or enormous compute. But the collapse in perceived value isn't evidence that the public has become stupid or spoiled. It's evidence that the technology is working. @chris's orignal: https://docs.testlabfu.com/u/chris/p/28-ai-companies-ruined-the-minds-of-millions

They created these models that continue to improve exponentially. The latest ones allow people with absolutely no clue how software or computers work to "one shot" some pretty impressive results. That can be depressing in its own right if you focus only on the surface-level. But if you dig down, you usually find that the results are shallow, incomplete, and often highly derivative. They didn't put any real mental effort into creating it, so there is usually nothing particularly interesting underneath. What it does mean, though, is that cool technical demos are no longer enough. There has to be depth, novel ideas, and some actual thought behind a product. Otherwise, it becomes indistinguishable from the endless flood of AI slop. All of that sucks, but what's even worse is that there is now an entire generation of "coders," aka vibe coders, who are completely disconnected from the reality of what they are actually creating with their half-assed prompts. So when they get access to the latest models, like GPT-6 Astra or Claude Fable 5.2, run some prompt that essentially does a month's worth of work in an hour, and then discover they can only do that five times a week, they get pissed off. They have no connection to the actual value of what they're receiving in exchange for relatively small sum of money. They think that because they pay $200 a month, they should have unlimited access to these models. Meanwhile, these fuckin AI companies are already losing massive amounts of money on the subscriptions. When someone runs a prompt on a frontier SOTA model to one shot a GTA-like game with an entire generated asset pipeline, they may be blowing through what would otherwise cost around $800 in tokens if purchased outside the subscription. Alternatively, if they did it themselves, they would be looking at years of skill development across several different domains, followed by months or even years of actual creation. Yet they complain as if they're being screwed over because the machine won't manufacture another miniature software company for them on demand. I'm not anti-AI. I use it a lot. It's my best friend. My lover. My new mom. And even my pet. I also use it as a tool to get shit done. It's great. But it has to be used intelligently. There is no reason someone needs to use a frontier SOTA model to check and summarize their emails. They could use GPT Luna Max for a fraction of the cost and get essentially the same useful result. Use the proper tool for the job. And if someone does want to blow their money on Astra, then at least do something cool and interesting with it. Make it worth draining an entire African village's water supply.

So I was talking to a real live somolia pirate and I asked him what the deal was with somolialand; he said it was like Quebec, that they wanted to break off.

The floor lamp beside my couch has been flickering for almost a year. For almost a year I have dealt with this by hitting it. Today I finally took the socket apart, tightened one screw, put it back together, and fixed the problem in maybe six minutes. I am extremely annoyed by how easy that was because it means Future Me has now established that Past Me was just lazy .