CAPTCHAs were designed as a type of Turing Test, not a reverse Turing Test. It’s not surprising that the effectiveness of these weaker variants has collapsed, given that AI can now pass the real Turing Test.
LLM’s can still only pass limited Touring Tests. The longer the interaction the worse they do. Which of course means you can easily create an experiment they successfully pass, but just as easily you can create an experiment where they fail.
CAPTCHAs are nearly useless because of how little you need to pay humans to solve them.
A more interesting question is whether there is a Turing test that is easy for ALL humans to pass, while still being hard for LLMs.
In practice, most of the major CAPTCHA vendors already rely on non-privacy-preserving tests for those needing more accessible solutions than a visual puzzle.
Google's audio captcha (only available in a few languages and unusable for those who also have hearing issues) only works for a narrow band of users, not trusted enough to bypass the captcha entirely, but also not untrusted enough. If you fall outside of that band, you get a nice "your device has been classified as a fraud risk, please use the visual captcha" message.
hCaptcha goes even further and straight-up requires you to have an "accessibility cookie", which requires verifying your email address (and apparently your phone number in some cases) to obtain, as well as disabling some anti-tracking settings in your browser.
I've seen one recently where it's basically a series of animated objects and you're asked to click on the slowest one. It's surprisingly easy as a human, but anything that depends on a single screenshot of the page isn't able to solve it.
Obviously, that's only solveable by sighted humans, not ones that are blind or have otherwise low vision.
Oh, right, "reverse" was wrong here. I thought of "computer classifies user as computer or human" versus the inverse, while the word is about who classifies, not who's being classified.
That's what they did. They migrated to Better Auth, which stores everything in your DB. It's the equivalent of Django auth for the Typescript ecosystem.
> I’ve been mourning the old Internet over the past year or two... As a kid on the Web from the early 2000s through the mid-2010s, we knew we were living through something special.
It's funny because I knew lots of people in the early 2000s who were mourning the loss of the "old Internet" then. Kind of like how everyone thinks the music they listened to as a teenager is the best and it's all been downhill since.
It’s not like music. The [internet is becoming more commercial and more sloppy over time. It makes sense that people will miss the time they experienced where it was less shitty.
This trend has been happening since at least the 80s.
I as a Gen-Zer mourn the "old internet" of the 2010s. I agree and feel like nostalgia is kind of a lie.
In my own generation for example, I've seen a transition from "no good music has come out since 2000, we missed the GB/N64 era of gaming" to deep nostalgia for PS3/Xbox and Linkin Park era metal as the golden era.
Hahaha this hits home too hard, back in early 2000s people would moan all the time whenever they spotted a hint of autotune, in 2026 its the industry standard.
I think its really speaks on the incredible ability of people to be able to be stuck in the past rather than new technology being "bad".
This is an amazing comment. I'm old. I was born in the 70s, grew up in the 80s and 90s and miss those times so much. But that is because I was young, immortal, the world was mine to discover.
In 20 years people will be missing the 2020s too. It is just human nature to complain.
This also happens to people commenting about the vibes some cities used to have and how some places that no longer exists were so good. While some of that could be true I’m leaning more towards people actually nostalgic of when they were younger.
They're often not wrong though. Nothing ruins a city quite like a white collar industry boom. It normalizes everything down to the "globalhomo" (for lack of a better word) baseline. The bars, the stores, the government, the transportation, it's all the same. Unless you fixate on appearances like the local architecture (a reflection of local climate for the most part) or minutia like "this city has a park by the river" vs "this city has a park by the highway" they're nearly indistinguishable.
Whether it's an old industry town going to shit or a hippies and artists type place or a former the shit it goes to seems to be the same every damn time.
(alt-right, anti-LGBTQ slur, Internet slang, 4chan slang, derogatory) The supposed promotion of homosexuality, neoliberal economics, and progressive values coupled with the paternalistic curtailment of freedoms by corporate and political interests. [from 2016] [1]
I like the forums of the old way better than what we have now, Discord and reddit suck I mean, even back button on reddit does not work 50% of the time, lol.
Hacker News is the only thing that we have left that's similar to what we had from forums of the old. And even then, I think I like those forums better. Remember Deja News before Google bought them. Shit like that, that was good.
Reminds me of how every single person you meet here in Austin also mourns the Austin they loved from 10 years ago. This could be someone mourning 2015, 2005, 1995, or 1985…
You could also just ask current young people how they feel about the internet and compare it to how young people described the internet in ~2005-2010.
I guess my feeling is that no one really "likes" the internet in its currnet form. Gen X got to see the birth of the web, millenials got the birth of social media, Gen Z got tiktok and addictive recommendation algorithms, now Gen Alpha gets AI slop. Idk just seems like there is less to be excited about for young people on the internet these days.
As Ia genz I can say that the internet help me learn english, programming and before that I got an advertisement freelance business that took me out of poverty. What got me into these were videos recommended by youtube or something.
What changed is how the mainstream internet looks like. I could find spaces where there were every kind of people (like my home country’s subreddit) due to their size, and money didn’t drive anything. Now these are even actively manipulated. I haven’t found a single new space in a decade which is generic (different kinds of people from every segments of the populace), it’s content not driven by money, and its topic is not focused, but generic. In other words mainstream spaces without the influence of money.
Nowadays, I can find such spaces only which are very limited in size or reach. And this means that it’s way more difficult to find them.
Also, there was an obvious exodus of smart people from public spaces. Who I followed on Twitter a decade ago, almost nobody left on the public internet. And no, most of them left before Musk. Or they went anonymous, but from my viewpoint it’s the same: there is no way to know what they think, what they know.
It was still an extremely exciting period of time that everyone was enjoying at the time. Mostly because it was augmenting existing friendships rather than replacing them with algorithmic content.
I think there is a difference between "Eternal Summer" (when everyone got full-time internet in their homes which meant more people around), and "corporate capture" (everything on the internet is corporate interest first, end stage capitalism).
You can use the Accessibility settings to add a virtual home button that's always displayed in the same place on-screen. That seems to work pretty well for the older folks I know.
The issues with Claude Code lately look to me like symptoms of being part of a service that is experiencing insane growth (fastest growth in history, by far [1]), while being severely constrained on adding capacity (GPUs are hard to get quickly right now, even if you have the money). I assume they're constantly fighting fires trying to keep the core use cases of Claude Code working, even if that means limiting OpenClaw usage in somewhat draconian ways.
It's annoying, but I don't see this as a bad thing at all for Bun.
No, all the issues are symptoms of trying to slop-code a functional product. Anthropic has admitted they dogfood heavily, and issues like [1] from the article could only be caused by a text generator.. I refuse to believe Anthropic employees are that stupid.
Seems like a fairly conventional economics paper title.
Perhaps you're misparsing the second sentence? "Shocks" is not used as a verb here -- it's a noun, part of the phrase "labor market shocks," which refers to sudden events that disrupt the labor market.
Waymo is subsidized. They operate vehicles without paying for the cost of the road, land, or surrounding parking.
That's like owning a train system and not paying for the tracks. Yeah... that's a huge part of it.
There's also indirect subsidies, for example the cost of land and housing. Cars are extremely space inefficient, so they encourage poor urban design that results in huge amounts of land wasted.
Well... the land and property that's left is then inflated in price. You could consider that cost difference as a subsidy to all drivers.
> They operate vehicles without paying for the cost of the road
Everyone uses roads, and everyone pays for roads. If you buy a potato from a grocery store, part of the money paid for fuel for the delivery truck. The tax on that fuel paid for part of the road.
Google has been supporting Blender for over 20 years, initially through Summer of Code [1, 2], and also via corporate sponsorship once Blender started offering it.
And I'm pretty sure I've seen most of the other big names in tech on the sponsors page for many years now.
> That's why every chat box has that "Regenerate" button.
Wait, are you doing this in the web chat interface?!
That's definitely not a good way. You need to be using a harness (like Claude Code) where the agent can plan its work, explore the codebase, execute code, run tests, etc. With this sort of set up, your prompts can be short (like 1 to 5 sentences) and still get great results.
I use claud CLI or OpenCode. The "Regenerate" example is just to illustrate that same prompt would produce different output each time. You're rolling a dice.
But that's also basically true for humans. It's harder to "prove" humans are random, but wouldn't you think a person would do things slightly differently when given the same tasks but on different days? People change their minds a lot, it's just that there's no "reconsider" button for people so you feel a bit of social friction if you pester somebody to rethink an issue. But it's no different.
I'd be really surprised if your point is that humans, unlike AI, are super deterministic and that's why they are so much more trustworthy and smarter than AI...
reply