evilduck

Most of the new Github bloat will just be thrown away. Vibe coding scratches an immediate itch and it's easy to do. Once the problem changes nobody is going to update the first project because that's hard, they'll just vibe code an entirely new solution leaving the first to rot until they delete their dead repo clutter or move on from the company and the account and all of its repos are deleted in one fell swoop.

llama was unintentionally released and has been a flash in the pan for them since. They've failed to capitalize on it at all post-v2. v3 models were a launch mess and quickly surpassed. The only thing it has left in 2026 is name recognition. Honestly their non-LLM things are more important and relevant, like their SAM models. https://ai.meta.com/resources/models-and-libraries/ but none of these seem to be making them serious money or exist and standalone products.

Seems like the entire field of alchemy and later chemistry was precluded by a "oh wow" or "that's interesting" or "help, I can't breath", not derivations from first principals. Throwing stuff together just to see what happens and then working backwards from interesting outcomes to recreate and understand those interesting outcomes seems like a perfectly valid if not chaotic approach for happening upon artificial consciousness.

Their gas is priced competitively. Everything is hard to compare because it's usually not even available aty other gas stations. It's like half of a Bass Pro Shops tacked on to a bizarre bbq restaurant that lacks seating. Comparing it to 7-11 or a Sinclair or Conoco gas station makes little sense. It's closer to a Love's Travel Center, and in which case their pricing holds up.

You can also build systems that require secondary approvals without needing approval escalation up the chain. Creeping on women is a lot less likely if you need a peer or even a subordinate to review what you're doing.

Yeah: It's a 128 year old home in a small town in the middle of nowhere. Hit it up in street view and look at the dilapidated neighborhood in the dying town it sits in, surrounded by a 50 mile radius of farm land.

Unlike ancient European masonry homes in some of the most passive of climates globally, a 128 year old wood construction house in the central United States is not desirable for most people. This one is in a location that gets hot summers and cold winters. Homes of this age come with substantial upkeep and modernization costs, they lack modern amenities like central air and heating, good insulation, or even a kitchen that fits normal modern appliances, they creak everywhere and are drafty, floors aren't level or are warping, they have a distinctive odor that's impossible to remove, and they harbor surprises from all sorts of things that the last 7 generations of owners did for themselves over the decades. My family has owned similar properties in Missouri. The price is reflective of the market demand, not the losses the owner dumped into it.

>the best ideas float up to Codex/Claude imo

They only float up if people create things like RTK and other people try them though.

It's fair to sit this one out and let others figure out if it's worth it or not but tools like RTK, Headroom, caveman mode and others do reduce input and output tokens that need to be processed, and for local LLMs that can have measurable speedups. Whether or not that ultimately hurts the resulting output I don't have enough data to say, but I am happy to play with them to find out.

Sure they will. When the real or perceived cost of addressing supply chain attacks exceeds the cost of changing tooling workflows, they will switch.

You're describing the FBI or your state level equivalent. And they actually do exactly what you are describing, but in measured efforts. I've even had them come by my place of employment before. They clearly lack the resources to work at this scale though.

The problem with a phone number you suggest is that it will get spammed and abused with fraudulent imposters too (the complete and utter destruction of trust in phone calls and text messages should also be corrected by the government, but that's a different topic).

https://www.fbi.gov/investigate/cyber

This is a real world trolley problem scenario. You can break workflows or you can let everyone get pwned by supply chain attacks. Which is the greater harm?

"Oops, we uploaded the wrong files" is the standard deflection every time people like this get caught.

Look up "Reflection 70B" drama.

I'll second this, with the caveat that I've not yet tried to build anything with Fable.

Every engineer can now produce things in the front end that doesn't look like complete and utter garbage, sure, but everyone is also producing the new-era of Twitter Bootstrap pages. It all has the same touched-by-AI look and it might as well be customer kryptonite from everything I've experienced at my workplace with customer surveys and collaboration. It has raised the floor substantially for internal tools and admin pages though.

The argument seems to be that they think LLMs made the implementation languages equal or irrelevant, therefore PHP is as good as any of them now. The flip side of this argument is that PHP is not any more compelling than it was before, it's just adequate now that so many deficiencies have been addressed.

No, I don't know. Maybe this exact reply is why you get the interactions you do though.

You perceive opposing viewpoints or poking holes in logic as "hating on you", which is playing the victim, followed by alluding to conspiratorial nonsense against you.

> and this is about not stepping on someone else.

This is asserting personhood or consciousness of LLMs by default in your phrasing and then warning me about the dangers of violating your assertion. You're making the same wager and mistake. There's no important difference, you have no evidence for LLMs being a "someone" any more than you do for a god existing. Warnings about made up things hold no weight.

Mostly because perverts tend not to press charges when they're confronted, not because they wouldn't have legal standing. Also your scenario is clearly not what I was replying to about civil disobedience.

I don't think you could plausibly do this and only catch a property crime charge. If you're caught, forcefully removing worn objects from another person will almost certainly catch you a misdemeanor battery charge in most US jurisdictions.

I'm no lawyer and things vary by location, but clothing is generally considered an extension of the person and usually touching their worn objects constitutes physical contact with the person themselves. Doing so with intent of committing criminal mischief, vandalism, or felony property damage will get all of them thrown at you. If you hastily do so and happen to harm the person in the process (since you're naturally grabbing at someone's eyes, that seems like a serious risk), there's a good chance you'll be given an aggravated or felony battery charge instead.

This seems like a variant of Pascal's wager.

Websites no, but there have been many Mac apps that I have paid for even though a lower quality free option existed.

I think we are approaching that now, with correct expectations. With frontier large models you can often one-shot tasks with vague prompts for stuff like creating CRUD APIs and dashboards around a simple data model since it's such a solved-problem now. With something like Qwen3.6 27B or 35B-A3B and a Strix Halo level computer or a MBP with 32GB or more or RAM, you may need to be more explicit and stay involved and be a little more patient, but you can absolutely get work done with it or delegate tasks to it successfully.

My Framework Desktop does a lot of similar work as my Claude subscription at work (Cowork, chats) for 100W of power draw and a little patience waiting for a slow GPU with limited memory bandwidth to crunch the numbers. Agentic coding is obviously weaker but CRUD development and visualization dashboards are within reach, and I'm usually pleasantly surprised at its ability to self-manage devops.

Poke looks like a startup with a 2 month head start that's also unvetted in the market, not a case study of permitted behavior and success at subverting iMessage with agents. Does anyone know if Apple even knows Poke exists yet?

Likewise, Poke also looks doomed. They're creating... OpenClaw but worse. OpenClaw hype and interest has recently fallen off a cliff, I don't think the angels will be getting their money back on that one.

How are your financial incentives aligned against sending spam? From this side, your words seem hollow and the typical viability of these businesses relies on sending spam.

The tourist portions of the city are created and supported by the surrounding population; they generate the parade, music, and celebratory culture year round. We can't just abandon or move the residential areas and keep the rest intact without it becoming a Disney main street facsimile version mocking what it once was, which would destroy the tourist appeal outright.

Wow, someone mentions fan speeds on their Thinkpad and you go into an unprompted dog whistle rant mode. A certain OEM lives rent free in a blighted mind.

Some people are on the bottom 5th percentile of the intelligence bell curve. Some people also enjoy having their balls being smashed in by high heels. Some people... blah blah blah. What a weasel statement. You are describing a straw man scenario that doesn't normally play out in the real world as such.

Replying with AI is the first obnoxious act. People reactive negatively to it should be expected.

You entirely missed the point in the article: Everybody has access to AI. Nobody needs another person proxying it slowly and poorly in a worse medium than direct access.

Replying with AI responses is equal to saying that you're no longer relevant or valuable to the discussion. If you are annoyed by that and want to block people, fine, obviously nothing of value will be lost because they can just go to the source directly next time instead of you.

They're not doing anyone any favors by disclosing it either, and anyone replying with AI verbatim but undisclosed also isn't going to be savvy enough to hide the other tells and quirks. Assuming they're ever asked again for their "input".

I'm baffled how you think syntax doesn't matter. Syntax affects how hard something is to understand when you read it, how hard it is to physically input, how hard it is to make mistakes, how hard it is to parse, interpret or compile.

Like most MS vs. Apple differences, it comes down to a matter of taste. They've added quite a few AI enhancements across their apps and operating system, but they are mostly feature enhancements and not major AI branded efforts. Having a Summarize button in Mail.app where it's contextually relevant or having text improvement menu options in text fields vs. slapping a major "Copilot" tab into everything.

Their use of AI so far has been much less "let AI take the wheel and brand it as a product itself" and more "use AI to improve an aspect of <user need>".

How many people from that era still exist at Apple to be holding a grudge like that? Genuinely curious, since it's been a very long time since he was last involved at Apple.

Living with laptops with touchpads for a couple of decades as a righty, I can use a touchpad perfectly fine with either hand, but a mouse is still right handed for me.