Nearly all of the Google images results for "baby peacock" are AI generated

As mentioned, product comparisons are a big one but another worrying area is anything medical related.

I was trying to find research about a medicine I'm taking this week and the already SEO infested results of 5 years ago have become immeasurably worse, with 100s of pages of GPT generated spam trying to attract your click.

I ended up ditching search alltogether and ended up finding a semi-relevant paper on the nih.gov and going through the citations manually to trying and find information.

It is quite surreal to witness. It is certainly fueled by the commercialization of internet due to ads and centralization to user hostile platforms.

The old internet seems to be doing much better. But it lost most of its users in the last 15 years...

People keep saying this and I keep warning them to be careful what they wish for. The most likely outcome is that "certification of human generated content" arrives in the form of remote attestation where you can't get on the internet unless you're on a device with a cryptographically sealed boot chain that prevents any untrusted code from running and also has your camera on to make sure you're a human. It won't be required by law, but no real websites will let you sign in without it, and any sites that don't use it will be overrun with junk.

I hate this future but it's looking increasingly inevitable.

https://x.com/samhenrigold/status/1843040235325964549

...or a sensical comparison where it just completely misses the point.

https://i.imgur.com/FotFZ3F.jpeg

Personally, I'm excited about more generative AI being added to my search results, and I'll probably switch to whichever search engine ends up with the best version of it.

Not only that, we desperately need cryptographic prof that content X was produced by person Y.

I wrote some ideas up about this many years ago: https://github.com/pjlsergeant/multimedia-trust-and-certific...

I'm talking about those 300-word, ad-ridden crap articles that are SEO'd right to the top, and if you're lucky you might get the 3-word answer you were looking for: "<300 words of shit>... and in conclusion, <1-step answer>.". Anyway, humans have been getting paid pennies to write those for a while.

AI just turns the throughput on that up to 11, where there's just no end in sight. I think this is like the primary failure mode of AI at this point. It's not going to kill us - we're going to use it to kill the internet. OTOH, maybe then we just go outside and play.

And long before the Internet, people slapped random concoctions together and sold them as medicine, advertising them as cure-alls.

I prefer text content to videos by a long shot, but genuine, human text content is almost dead. Reddit might be one of the rare exceptions for now. There are also random, still active, old school forums for lots of things but they tend to become extremely hard to find.

For example, https://archive.org/details/stanleyhomerepai0000fine/page/14... links to the chapter "Painting Trim the Right Way" from the book Stanley Home Repairs, 2014.

Could also look at used book stores. Home repair hasn't changed much.

Edit: Could even fire up Wine and try the CD-ROM "Black & Decker Everyday Home Repairs" (published by Broderbund) at https://archive.org/details/BlackDeckerEverydayHomeRepairsBr... . https://www.goodreads.com/book/show/3424503-everyday-home-re... says;

> Like its predecessor in book format, the CD-ROM version offers easy-to-follow, step-by-step instructions on more than 100 common household problems, from how to fix a leaky faucet to repairing hardwood floors. What's more, the CD-ROM version incorporates animation and narration to help make the repair project even easier to understand and complete. Instructions can be viewed one step at a time or all at once, and, if desired, can be printed out and taken directly to the repair site. Included with each repair project is the projected time needed to complete the work, estimated cost, and a list of materials and tools needed.

That sounds pretty nifty, actually!

All of the flat white MDF trim you buy is primed and ready for painting, too.

The information on working on my new car is non-existent other than Youtube videos where the majority is just a random dude who knows nothing filming himself doing a horrible self repair.

This is frightening and, I fear, true.

But I'd also add one odd little counterpoint: some of the most useful discussions and learning experiences I've had in the last four years have happened in private Facebook groups. As soon as the incentive to build a following using growth-hacking and AI -- which private groups mitigate to a greater extent -- is taken away, you get back to the helpful stuff.

The FreeCAD group on Facebook is great, for example. And there are private photography groups, 3D printing groups, music groups etc., where people have an incentive to be authentic.

Public Facebook feeds are drowning in AI slop. But people who manage their own groups are keeping the spirit alive. It's almost at the point where I think Facebook will ultimately morph into a paid groups platform.

I see a lot of people talk nostalgically about blogs, but they were an early example of the internet changing from ever green content to churning out articles on content farms. If people remember the early internet, it was more like browsing a library. You weren’t expecting most sites to get updated on a daily - or often even a monthly - basis. Articles were almost always organized by content, not by how recent they were.

Blogging’s hyper-focus on what’s new really changed a lot of that, and many sites got noticeably worse as they switched from focusing on growing a library of evergreen content to focusing on churning out new hits. Online discussions went through a similar process when they changed from forums to Reddit/HN style upvoting. I still have discussions on old forums that are over a decade old. After a few hours on Reddit or HN, the posts drop off the page and all discussion dies.

That's retweets.

> undesirable human-generated content was already starting to intrude into this business model, but now generative tools are also producing undesirable content faster than moderation can keep up.

> people will become disinterested, and tools which could previously use algorithmic heuristics to determine which content is good vs bad will begin to become useless.

So what these parts are saying is, tiny monoculture of bored college kids are always going to figure out the algorithm and dominate the platform with porn and spams and chew up all resources, and that both improved toolings and tie-in to monetary incentives intended to empower weaker groups to curb kids only worsens the gap, and that that's problematic because financial influencers are paying to be validated by the masses, not to be humiliated by few content market influencers.

But what is the problem with that? Those "undesirable content" producers are just optimizing for what market values most. If that's problematic, then the existence of the market itself is the problem. What are we going to do with that? Simply destroying it might make sense, perhaps.

hasn't publishing since Gutenberg been driven by the monetization of content? Looking at the history of the Catholic Church, potentially before that too.

This is the kicker. When unfettered by regulation or leaders/workers with morals, most industries would rather avoid human curation because they want to sell you something. Amazon sellers would rather you not see or not trust the ratings because they want you to buy their stuff without knowing it's going to fall apart. Amazon makes a profit off it, so they somewhat encourage it (although they also have the dual pressure of knowing that if people distrust Amazon enough they'll leave and go somewhere else, so they have to keep customers somewhat happy).

No, curation has to come from individuals, grassroots organizations, and/or companies without a financial interest in the things being curated - and it has to revolve around a web of trust, because as Reddit has shown, anonymous curation doesn't work once the borderline criminal content marketers find the forum and exploit it.

> The only way out I can see is something along the lines of human curation of human-generated content.

...however, unfortunately, curation doesn't solve the problem of people desiring AI-generated content. That's a much harder problem. Even verifying that something was created by a human in the first place is hard. I don't want to think about that. I'm just going to focus on curation because that's easier and it's also incredibly important for the lowering quality of physical goods as well.

There is only one way I can see things changing and people aren't going to like it. All content on the internet gets linked to a legal ID. Every post on facebook, every comment can be attributed to a real person.

I don't know what the solution here is. My guess is that the "public internet" will become less and less relevant over time as it becomes overrun by low-quality content, and a lot of communication will move to smaller communities that rely heavily verifying human identity and credentials.

The same could be said for regular search. Pretty much anything I search for yields a page of ads followed by pages of content farmers followed by pages of "almost sounds like experts but is still just a content farmer."

Financing the Internet with advertising has really made it difficult to find good quality content. The incentives are completely misaligned, unless you are a 'content creator' or Google.

The question is what comes next, and I don’t think anyone has an answer to that.

Source: every other piece of news or social media on the planet.

Would you pay money to Google or some other company in exchange for a genuinely good search service that prioritizes well written content, and avoids AI (or human) generated crapticles?

I'd like to think that too, but I wonder how long - if at all - this will be true. I "want" to like human generated content more, but I suspect AI may be able to optimize for human engagement more, especially for simple dopamine inducing content (like tiktok videos). After all, we're less complicated than we like to think.

>It’s crucial for AI tools to include robust detection mechanisms, such as reliable watermarks, to help other platforms identify AI-generated content.

This will never work, unfortunately. There's no way to exclude rogue actors, and there's plenty of profit in AIs pretending to be human. If anything, we will have to watermark/sign human generated content.

It's not helpful that you're making a binary distinction here.

As an example, as much as 10 years ago, I would find Youtube videos where the narration was entirely TTS. The creators didn't want to use their own voice, and so they wrote the script, and fed it into a TTS system. As you can expect from the state of the art at the time, it sounded terrible. Yet people enjoyed the videos and they had high view counts.

Are we calling this AI-generated?

We now have better TTS (without generative AI). Way better. I presume those types of videos are now better for me to watch. You may still be able to tell it's not a human because the tone doesn't have much variance. You'd probably have to listen for a minute or longer to discern that, though.

Are we calling this AI-generated?

Now with generative AI, we have voices that perhaps you won't be able to identify as AI. But it's all good as long as a human wrote the script, right?

Are we calling this AI-generated?

Finally, take the same video. The creator writes the script, but feels he's not a good writer (or English is not his native tongue, and he likely has lots of grammatical errors). So he passes his script to GPT and asks it to rewrite it - and not just fix grammatical errors but have it improve it, with some end goal in mind ("This will be the script for a popular video...") He then reviews that the essence he was trying to capture was conveyed, and goes ahead with the voice generation.

Is this AI-generated?

To me, all of these are fine, and not in any way inferior to one with a completely human workflow. As long as the creator is a human, and he feels it is conveying what he needed to convey.

I would love to take a first draft of a blog post, send it to GPT, and have it write it for me. The reason I don't is that so far, whatever it produces doesn't have my "voice". It may capture what I meant to say, but the writing style is completely different from mine. If I could get GPT/Claude to mimic my style more, I'd absolutely run with it. Almost no one likes endless editing - especially writers!

It's disappointing to see scammers and black-hat SEOs already leveraging Notebook LM to mass-produce fake podcasts and distribute them across various platforms.

Didn't they do something like that with AMP. I recall that if you were using chrome and visited an AMP site from Google the address bar would say site.com even though the content was being served from google.com.

Whether they actually do this (and whether there's any incentive to do so), is obviously not a given

https://wikipedia.org/wiki/Low-background_steel

So the pessimist in me can see the Internet being affected by the free-vs-premium formula: "basic" Internet with ads, tracking, AI fillers, limited access to +18 content, in the worst form comes with these pre-defined sites and "premium" that's free of these limitations but it also in time tries to squeeze more money from users - like "premium but with ads"

While I generally agree with your whole comment, I feel like this part has been true for years on social media well before AI generated content hit the scene.

Wikimedia is a fantastic resource.

The masses are fully here now. They're too passive to know or care what's going on. They stick with the path of least resistance: Google, Amazon, Reddit, Twitter, etc. No matter how hostile or shitty those options become.

We have to put aside the way we've thought about the internet before now because it doesn't apply anymore. There will be no more MySpace -> Facebook. The internet is no longer made up of a high enough percentage of conscientious and deliberate users to make a difference.

...

"Asynchronous, symmetrically anonymized, moderated open-cry repute auction. Don't even bother trying to parse that. The acronym is pre-Reconstitution. There hasn't been a true asamocra for 3600 years. Instead we do other things that serve the same purpose and we call them by the old name. In most cases, it takes a few days for a provably irreversible phase transition to occur in the reputon glass - never mind - and another day after that to make sure you aren't just being spoofed by ephemeral stochastic nucleation."

Fantastic book. I read it twice so far, highly recommended. So many little off-handed conceptual gems everywhere.

You talk about being a part of something "real and visceral" but you're complaining about the demise of being able to sit at your desk and see pictures of wildlife. Maybe it's okay that google image search dies and makes people go out and find the wildlife they want to see.

The internet, even in its best format (e.g. ad-free, free access information for all; and communication with all of humanity) has a ton of real downsides. It's not clear to me that AI should be strangled in its infancy to save the internet (which does _not_ exist in that "best" format).

- I am a thoughtful technologist, building real things for real people, concerned about others and the social impact of my work;

- they are greedy and ignorant, destroying society for short-term personal gain, no matter what the consequences.

It's human nature to put badness on an abstract them, but we don't get anywhere that way. It's good for getting agreement (e.g. upvotes), because we all put ourselves in that sweet I bucket and participate in the down-with-them feeling. But it only leads to more of what everyone decries.

You could say what you say about anyone at any time. Where do you draw the line? I guarantee you'll be guilty of the exact same thing. I don't want to generalize, but IMO this sentiment of yours, I hear most loudly from software engineers far removed from ordinary non-technical end users: is making beautiful new LISPs and CNIs and Python package auditing tools the only valid work with seemingly no tradeoffs?

That said, I don't really think this is a tide any individual market actor can reasonably stem. It's going to require some pretty fundamental changes in the way we use the internet.

Sick and tired of giving parasites benefit of the doubt they've long sucked dry.

(also noteworthy for the 'Publication' section near the bottom)

This is already impossible because it's impossible to enforce. You can't stop something running on a random laptop, and you can't stop models running on server farms in, say, North Korea.