News for hackers. Stuff that matters.
Re:Chatbot Hallucinations Are Poisoning Web Search by pseudosavant on Thursday October 05, 2023 @03:03PM (#37782738)
I wonder if there will be a human information/knowledge equivalent of low-background steel (pre-WWII/nukes). Data from before a certain point won't be 'contaminated' with LLM stuff, but it'll be everywhere after that.

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

Re:Chatbot Hallucinations Are Poisoning Web Search by DayDollar on Thursday October 05, 2023 @03:30PM (#37783161)
There will be a web of trust, with a valuation of nodes by trustworthyness. And people will get only one id for this. Ones name is ones value and a reputation will be a hard earned thing again.
Re:Chatbot Hallucinations Are Poisoning Web Search by ratg13 on Thursday October 05, 2023 @03:37PM (#37783252)
This was how the "internet" functions in the book "Ender's Game".

There is a small sub-plot about how he had to give a fake persona credibility on the untrusted network in order to be able to leverage a creating a fake account on the trusted network.

Re:Chatbot Hallucinations Are Poisoning Web Search by dredmorbius on Thursday October 05, 2023 @04:01PM (#37783555)
I find the xkcd interpretation more realistic: <https://xkcd.com/635/>

Explained: <https://www.explainxkcd.com/wiki/index.php/635:_Locke_and_De...>

Re:Chatbot Hallucinations Are Poisoning Web Search by notahacker on Thursday October 05, 2023 @04:42PM (#37784034)
I love that interpretation, but in today's retweet driven world of politically commentary, I actually find it quite plausible that pseudonymous kids with no grasp of the real world who think rational political debate is the nonsensical slogans they're spouting on the internet become major Twitter influencers that actual politicians want to court for their "authenticity" and "willingness to say the unsayable", and maybe their dank memes.
Re:Chatbot Hallucinations Are Poisoning Web Search by dredmorbius on Thursday October 05, 2023 @06:08PM (#37784904)
The conceit of Ender's Game was that thoughtful discourse would be influential online.

Reality has largely demonstrated that far more thoughtless propaganda of the Big Lie, Firehose of Bullshit (or Falsehood), associated with Russia, floods of irrelevance which tend to bury more significant stories, favoured by China, and outrage / hot-button topics, which are common in US-centric media, though a timeless technique.

Memes and simple messages attract attention and spread. Complex narratives and analyses ... not so much.

But yes, voices that deserve no attention whatsover have dominated the media landscape of the past decade or so. Not that this is entirely novel.

Re:Chatbot Hallucinations Are Poisoning Web Search by acover on Thursday October 05, 2023 @07:15PM (#37785415)
Good and unlikely predictions can thrust you into popularity: see deepfuckingvalue.
Re:Chatbot Hallucinations Are Poisoning Web Search by carlosjobim on Thursday October 05, 2023 @05:07PM (#37784301)
Isn't this how it has been since the dawn of time?
Re:Chatbot Hallucinations Are Poisoning Web Search by thenickdude on Thursday October 05, 2023 @07:24PM (#37785485)
Will you join my Webring?
Re:Chatbot Hallucinations Are Poisoning Web Search by grotorea on Friday October 06, 2023 @09:54AM (#37790832)
Not disagreeing, but that's the end of anonymity.

But yeah, maybe the idea that you can even 1% trust random content on the Internet without having a source doesn't really make sense if you think about it IMHO. Either you do this web of trust, coming from a well know real world source, or be Wikipedia-like with linked reliable sources for the viewer to check.

By the way, wasn't this how Google ranked pages back in the day? Ranking pages that get linked to higher? And even before that there were P2P web rings.

Re:Chatbot Hallucinations Are Poisoning Web Search by RandomWorker on Thursday October 05, 2023 @03:14PM (#37782920)
My sense is to avoid this have a personal blog.

That being said how many people write blogs with grammerly or chatgpt these days. The temptation to use these technologies all the time is too strong for even self preservation of your own (writers) voice.

My sense is that you use this technology you might be happy with the results at first but on later review you just notice something off in some sentences and maybe it just doesn’t flow right. I’m not convinced that it will replace writers jobs yet. Especially when you want to create something authentic and unique.

Re:Chatbot Hallucinations Are Poisoning Web Search by tredre3 on Thursday October 05, 2023 @03:29PM (#37783142)
> The temptation to use these technologies all the time is too strong for even self preservation of your own (writers) voice.

I don't know about that. I have played with ChatGPT/Copilot/etc enough to know what they're capable of doing. But the thing is, I enjoy programming. I enjoy breaking down a problem and solving it with code. I enjoy crafting elegant code. So I don't use AI even though I'm fully aware it could save me hours on projects. Why? Because I enjoy those hours very much.

Why am I telling you all this? Because I suspect many writers are the same and personal blogs are their canvas. They enjoy communicating. They enjoy crafting articles. They might have AI proof-read them, but they won't let them write everything. So, to me, there is hope that personal blogs will maintain their human element, as opposed to news websites or tabloids or learning platforms.

Re:Chatbot Hallucinations Are Poisoning Web Search by SoftTalker on Thursday October 05, 2023 @03:39PM (#37783280)
Agree. I've never even looked at any of these AI tools. I enjoy the process and the challenge of programming, and the rewards of doing it well. I have no desire for someone or something else to write code for me.
Re:Chatbot Hallucinations Are Poisoning Web Search by steelframe on Thursday October 05, 2023 @04:01PM (#37783564)
> So I don't use AI even though I'm fully aware it could save me hours on projects.

Enjoy this luxury while it lasts. Based on what I have seen in performance review committees for software developers, your peers who drive results faster than you do because they use AI will be rewarded more and will be more likely to survive rounds of layoffs when they inevitably happen.

Re:Chatbot Hallucinations Are Poisoning Web Search by JohnFen on Thursday October 05, 2023 @06:46PM (#37785214)
That's fine. I genuinely wouldn't want to continue working in an industry that worked like that anyway, so I'd just quit and keep on programming with my own projects. So that luxury will last as long as I want it to.
Re:Chatbot Hallucinations Are Poisoning Web Search by booleandilemma on Thursday October 05, 2023 @07:03PM (#37785331)
I've come to the same conclusion. I became a programmer because in school I loved programming and saw myself loving the job, and I do.

If the job changes that drastically I'll just have to quit and find something else.

Re:Chatbot Hallucinations Are Poisoning Web Search by EFreethought on Friday October 06, 2023 @04:32AM (#37788367)
But what if you have to use an LLM at your new job?
Re:Chatbot Hallucinations Are Poisoning Web Search by pseudosavant on Thursday October 05, 2023 @03:26PM (#37783109)
Sometimes the value is specifically because my voice won't come through. When I'm stressed and being asked for unreasonable things at work, I know that I tend toward passive aggression. But professionally, that isn't the way I want my message to come across.

I use ChatGPT all the time to suggest how I could make sure something isn't passive aggressive. It'll point out parts that aggression and suggested changes. It can be for a short slack message, or a many paragraph message.

Re:Chatbot Hallucinations Are Poisoning Web Search by floren on Thursday October 05, 2023 @03:47PM (#37783374)
I have definitely read "blogs" written by stitching together LLM outputs. For years people were advised that a technical blog "looks good on a resume" so we saw lots of lightly rewritten Stackoverflow content. Now it's gotten easier.
Re:Chatbot Hallucinations Are Poisoning Web Search by datadrivenangel on Thursday October 05, 2023 @03:21PM (#37783019)
There's the branch of philosophy called epistemology.
Re:Chatbot Hallucinations Are Poisoning Web Search by BitwiseFool on Thursday October 05, 2023 @03:46PM (#37783366)
Those simple web 1.0 sites made by college professors are a gold-standard in my book. I always enjoy finding them in search results. Although they are becoming increasingly rare.
Re:Chatbot Hallucinations Are Poisoning Web Search by dredmorbius on Thursday October 05, 2023 @03:58PM (#37783520)
Unfortunately, that's a trivial signal to emulate.

At a minimum, you'd have to validate them by confirming existence in the Wayback Machine.

Otherwise agreed that those are indeed high-signal documents. Increasing reliance on integrated educational software means that even such things as online syllabi are increasingly rare.

Re:Chatbot Hallucinations Are Poisoning Web Search by LordDragonfang on Thursday October 05, 2023 @04:57PM (#37784196)
The type of sites GP is talking about are typically hosted on .edu servers, under faculty webhosting (often featuring a "/~profname/" in the url). That's a non-trivial signal.
Re:Chatbot Hallucinations Are Poisoning Web Search by BitwiseFool on Thursday October 05, 2023 @05:30PM (#37784532)
Yes, I should have included that detail.
Re:Chatbot Hallucinations Are Poisoning Web Search by dredmorbius on Thursday October 05, 2023 @06:01PM (#37784841)
~/name at an edu is pretty attainable.

.edu domains can be had for any otherwise eligible "U.S.-based postsecondary institutions" per Educause: <https://net.educause.edu/eligibility.htm>

Pages at extant domains might variously be available to undergraduate or graduate students, faculty, staff, and adjuncts. Those might either directly host emulative material or be convinces or compromised into hosting content.

If there's one thing that the Internet's history to date has proved, its that perverse incentives lead to perverse consequences.

Re:Chatbot Hallucinations Are Poisoning Web Search by l33t7332273 on Thursday October 05, 2023 @06:48PM (#37785226)
It is not easy for a regular person to obtain access to a .edu webpage.
Re:Chatbot Hallucinations Are Poisoning Web Search by dredmorbius on Thursday October 05, 2023 @07:09PM (#37785378)
A "regular person" can:

- Enroll or be hired at an eligible institution. There are literally thousands of these.

- Bribe or compromise someone enrolled or hired at an eligible institution.

- Create a de novo eligible institution. For-profit colleges are not uncommon.

Someone motivated by profit or advantage would likely find virtually any of these options quite straightforward.

I'm ... somewhat pained that this needs to be spelled out.

Re:Chatbot Hallucinations Are Poisoning Web Search by l33t7332273 on Thursday October 05, 2023 @08:46PM (#37786031)
> Enroll or be hired at an eligible institution. There are literally thousands of these

You don’t generally get the kind of personal website being discussed to my understanding.

> Bribe or compromise someone enrolled or hired at an eligible institution

Finding a professor willing to stake their job and reputation for such a blatantly immoral scam seems hard.

> Create a de novo eligible institution

Is it actually easy for a regular person to create their own college?

I find the snarky finish to your comment obnoxious.

Re:Chatbot Hallucinations Are Poisoning Web Search by 90-00-09 on Friday October 06, 2023 @12:22AM (#37787096)
A "regular person" also can:

- start their own country and call it Edunistan

- bribe ICANN to take over the .edu TLD

- open a university in the new country

- spend 15 years earning a PhD at that university

- reserve ~/name and start posting LLM generated content

Re:Chatbot Hallucinations Are Poisoning Web Search by heavyset_go on Thursday October 05, 2023 @04:00PM (#37783541)
Can't prove it, but it seems to me like black text on white background sites from the past are poorly ranked compared to sites with "modern" layouts.
Re:Chatbot Hallucinations Are Poisoning Web Search by hashtag-til on Thursday October 05, 2023 @05:57PM (#37784794)
Yes. I love black text on white background. A rare find these days.

Browsing today is like: “You ask for a spaghetti recipe and the page tell you the whole history of civilization.”

Re:Chatbot Hallucinations Are Poisoning Web Search by zeroonetwothree on Thursday October 05, 2023 @06:02PM (#37784858)
Thats specific to recipes because they can’t be copyrighted
Re:Chatbot Hallucinations Are Poisoning Web Search by hashtag-til on Thursday October 05, 2023 @06:41PM (#37785169)
I had a look and definitely learned something today so #til.

Also, note to self to collect my favourite recipes in markdown files from now on.

Re:Chatbot Hallucinations Are Poisoning Web Search by xkcd-sucks on Thursday October 05, 2023 @08:13PM (#37785838)
Re:Chatbot Hallucinations Are Poisoning Web Search by grotorea on Friday October 06, 2023 @09:50AM (#37790779)
I heard that before and have trouble believing this is the cause at least for Internet recipes. Sure for a recipe book in 1950, but are recipe content farms going to sue each other? Isn't the lawyer costs way more than could be gained?
Re:Chatbot Hallucinations Are Poisoning Web Search by StableAlkyne on Thursday October 05, 2023 @08:26PM (#37785917)
Don't forget how Google will now drop search terms if it thinks you mean something else, or add unrelated synonyms to your search (presumably to "help" folks who aren't good at writing queries)
Re:Chatbot Hallucinations Are Poisoning Web Search by p_l on Friday October 06, 2023 @04:42AM (#37788404)
At least it still lets you force whatever you write in...

Allegro (big polish auction/e-commerce site) in their mobile app will unconditionally rewrite the search terms instead of showing you no results

Re:Chatbot Hallucinations Are Poisoning Web Search by HappyDaoDude on Thursday October 05, 2023 @07:02PM (#37785323)
I have a website that is just a few black text on white HTML files I maintain in whatever text editor I have at hand. Loads lightning fast, and if you cannot view it, its not a web browser. Last I checked, the total site size was about 60KB.

As time goes on, even the amount of text I am putting out get trimmed down. Make the words count, don't count the words.

Re:Chatbot Hallucinations Are Poisoning Web Search by gorwell on Thursday October 05, 2023 @07:24PM (#37785498)
Further, is it true that Google factors in whether sites have ads?
Re:Chatbot Hallucinations Are Poisoning Web Search by semi on Friday October 06, 2023 @11:51AM (#37792384)
They're even better than black text on white backgrounds. They're unstyled and use your browser default styling. granted it's rare anyone configured those so it's almost always black on white.. but for people who do specify their own preferences it's really nice to have them respected and not rely on hacking in my own css or js to override theirs
Re:Chatbot Hallucinations Are Poisoning Web Search by MrVandemar on Thursday October 05, 2023 @06:29PM (#37785077)
search.marginalia.nu is a great place to find those sites, and some more interesting stuff besides.
Re:Chatbot Hallucinations Are Poisoning Web Search by StableAlkyne on Thursday October 05, 2023 @08:24PM (#37785909)
Paul's Notes remains the absolute best textbook Calculus and intro to DiffEQ! Not sure if it's been updated since 2005, but I mean, it's not like they're discovering new Calc II methods! Being that it's not a $200 textbook rehashing the same stuff as the last 10 editions, it's easily one of my favorite websites
Re:Chatbot Hallucinations Are Poisoning Web Search by ryanklee on Thursday October 05, 2023 @04:23PM (#37783830)
People are vastly everestimating how unique this problem of hallucinations is.

It seems to me it relies mostly on discounting just how much we've already had to deal with this same problem in humans over the millenia.

The problem of proliferation of bad information might be getting worse, but this isn't native to generative AI. The entire informational ecosystem has to deal with this. GPTs compound the issue, but as far as I can tell, no where near what social media has forced us to deal with.

Re:Chatbot Hallucinations Are Poisoning Web Search by BobaFloutist on Thursday October 05, 2023 @04:37PM (#37783987)
The thing is when you call a human on bullshit, they usually can't back it up well enough to pass the smell test. When you call an AI on bullshit it can instantly fabricate plausible, credible seeming sources/evidence.

A human's lie is different than an AI's hallucination, since it's still based on (distorting) the truth, whereas the hallucination is based on an invented reality (yes I know it's applied statistics and there's no true model of the world in there, but it can report as if there is)

Re:Chatbot Hallucinations Are Poisoning Web Search by ryanklee on Thursday October 05, 2023 @04:57PM (#37784203)
Intelligent people can fill the void of ignorance with plausible sounding but factually incorrect information. They are apt to engage cognitive biases in such a way that the biases produce assertions that are deeply indistinquishable from factual assertions. They fool themselves in this way and they fool others. This happens all the time.

LLMs are no different in this respect.

Re:Chatbot Hallucinations Are Poisoning Web Search by TerrifiedMouse on Thursday October 05, 2023 @10:02PM (#37786450)
Sure people lie and spread misinformation (willingly or unwittingly). But coming up with plausible sounding lies take effort.

LLMs on the other hand are amazing and prolific liars and can produce a lot of bullshit for a price that’s effectively free - in fact it’s cheaper to create a LLM that’s inaccurate than one that’s … less inaccurate. The truth to lie ratio on the internet is about to take a huge hit. LLMs really are a Pandora’s Box.

Re:Chatbot Hallucinations Are Poisoning Web Search by gyudin on Thursday October 05, 2023 @05:59PM (#37784822)
It’s not a big deal, there are many ways to handle it. It just has some overhead costs. LLMs that are offered to general public are more of a POC and they are making sure to use as little resources as possible.
Re:Chatbot Hallucinations Are Poisoning Web Search by 127 on Friday October 06, 2023 @03:23AM (#37788034)
Evidence can easily be cross referenced and if the link LLM gives is dead it means it's lying.
Re:Chatbot Hallucinations Are Poisoning Web Search by cscurmudgeon on Thursday October 05, 2023 @06:09PM (#37784906)
How do we know you are not hallucinating this comment?
Re:Chatbot Hallucinations Are Poisoning Web Search by wellthisisgreat on Thursday October 05, 2023 @06:15PM (#37784960)
Yeah if you think about it, there is no history for example, as all we have in that domain is just someone’s perspective on some events. They may or may not have agenda but that’s beside the point.

That soft data could have never been trusted, rhe information that can be verified (calculations etc.) seems safe from LLM

Re:Chatbot Hallucinations Are Poisoning Web Search by blibble on Thursday October 05, 2023 @06:44PM (#37785200)
humans can only produce semi-convincing bullshit at a limited rate

with AI this limit is all but removed

all the human generated bullshit ever created will soon be dwarfed by what AI can vomit out in an hour

Re:Chatbot Hallucinations Are Poisoning Web Search by HappyDaoDude on Thursday October 05, 2023 @06:54PM (#37785263)
Like most things in the world. The problem isn't necessarily the technology but the scale at which it is implemented.
Re:Chatbot Hallucinations Are Poisoning Web Search by darkerside on Thursday October 05, 2023 @09:11PM (#37786171)
How do you know that? A human that produced fully convincing bullshit would be definitionally undetectable to you
Re:Chatbot Hallucinations Are Poisoning Web Search by TerrifiedMouse on Thursday October 05, 2023 @10:13PM (#37786503)
Because everyone has at least tried to lie - not necessarily for malicious reasons, e.g. white lies - a few times in their lives and know it’s not easy coming up with plausible sounding lies. It takes effort. Sure it might be effortless for some psychopaths but there are a limited number of such people. Overall the internet was surprisingly usable despite all the liars - even now with SEO and content farms; I think even SEOs realize that it’s better to provide information that’s true than lie if it cost them the same.

LLMs changes all that. They can produce content on a massive scale that can drown out everything else. It doesn’t help the fact that more inaccurate LLMs are cheaper and easier to create and run - think about all the ChatGPT3 level LLMs vs ChatGPT4 level LLMs; guess which ones SEOs will gravitate towards.

Re:Chatbot Hallucinations Are Poisoning Web Search by darkerside on Friday October 06, 2023 @03:54PM (#37795448)
Again...

> Because everyone has at least tried to lie - not necessarily for malicious reasons, e.g. white lies - a few times in their lives and know it’s not easy coming up with plausible sounding lies.

How do you know that?

Re:Chatbot Hallucinations Are Poisoning Web Search by TerrifiedMouse on Friday October 06, 2023 @07:11PM (#37797496)
> How do you know that?

Common knowledge.

Regardless, the amount of false information on the internet used to be limited by the amount of people producing it. Some of it is deliberately produced misinformation. Some of it are just mistakes due to carelessness, others due to willful negligence - example of the latter would be content farms who make their money off flooding search engine result pages and pushing ads in your face when you visit their website; who couldn’t care less what if you got what you came for.

Despite all that the internet was still useful. There are enough people posting accurate information such that the signal to noise ratio is good enough.

LLMs will change all that. They have the potential to flood the internet with hallucinated falsehood. And because bad LLMs are cheaper and easy to create and operate, those will be the majority of LLMs used by aforementioned content farms.

Re:Chatbot Hallucinations Are Poisoning Web Search by darkerside on Saturday October 07, 2023 @12:28PM (#37803082)
I guess I just disagree on either how capable people are of thinking on their feet, or what your definition of bullshit is. Still not sure which.
Re:Chatbot Hallucinations Are Poisoning Web Search by TerrifiedMouse on Saturday October 07, 2023 @07:20PM (#37806350)
Well, you can be sure the SEO creating content farms will pick the path of least resistance.

They won't make up a lie when telling the truth is far easier and when they do have to lie, it's a slight bit more effort.

But that's a moot point with LLMs ...

If/When they use LLMs, they will pick the cheapest LLM possible which will produce the most garbage - they might not even bother keeping it up to date; why bother when hallucinated BS sound just as convincing.

It has already started to happen.

Re:Chatbot Hallucinations Are Poisoning Web Search by cleandreams on Thursday October 05, 2023 @07:02PM (#37785329)
Uh, no. Generative AI does not have a standard of truth. It generates text according to the probabilities given what it learned from its data set. There is no systematic modelling of a world that it can test its assertions against. What are called hallucinations are a fundamental property of the approach. The hallucinations are probable -- just not true. The model has succeeded but the assertions are false. This has to be understood or the models will mess stuff up. A lot of checking is required.

Humans always assemble information according to a standard of truth. It is a big part of how humans learn. No method is perfect but the human method results in fewer routine hallucinations.

Re:Chatbot Hallucinations Are Poisoning Web Search by pixl97 on Thursday October 05, 2023 @07:57PM (#37785725)
A standard of truth is doing a lot of heavy lifting. Myself I would say we assemble information based on the limitations of the body we exist in. For example gravity is important to us because if we disobey its laws we may very well die.

Embodiment and multi modal AI will likely provide such filters or limits to AI in which it can derive truth.

Re:Chatbot Hallucinations Are Poisoning Web Search by ryanklee on Thursday October 05, 2023 @08:01PM (#37785748)
> the human method results in fewer routine hallucinations.

I'd love you to produce data to back this up.

My guess is that you are wrong, on the basis of how often I discover that I'm full of shit and how often I discover other people are full of shit.

Humans are built for being wrong just as much as being right. We wouldn't have such complicated institutions and social structures built around controlling for those symmetric capacities if it weren't the case.

Even then, we find ourselves surrounded and overcome by falsehoods of our own design.

Re:Chatbot Hallucinations Are Poisoning Web Search by wavemode on Thursday October 05, 2023 @09:18PM (#37786210)
You could probably just look up data on mental illness and/or pathological lying? Most people will say "I don't really know the details of that" rather than write you an essay of seemingly plausible but completely made up nonsense. Not all people, sure, but most.
Re:Chatbot Hallucinations Are Poisoning Web Search by ryanklee on Friday October 06, 2023 @10:33AM (#37791316)
The kind of incorrect information I am referencing is not pathological, but the type that is generated by cognitive bias and woven into the functional fabric of everyday life. It is pervasive and most people don't notice it (even when aware of specific bias types).
Re:Chatbot Hallucinations Are Poisoning Web Search by gorwell on Thursday October 05, 2023 @07:32PM (#37785556)
If there's an upside here, it's that humans will now be forced to refine their BS detectors.

In doing so the species will improve critical thinking skills which can be applied to all information regardless of source. Which, I agree, was often BS to begin with and was often able to skirt on by without notice.

Re:Chatbot Hallucinations Are Poisoning Web Search by mvdtnz on Thursday October 05, 2023 @07:50PM (#37785681)
Unfortunately that didn't happen with other innovations of misinformation at scale (such as social media) so I'm not so optimistic.
Re:Chatbot Hallucinations Are Poisoning Web Search by mvdtnz on Thursday October 05, 2023 @07:48PM (#37785668)
Not one single solitary soul has ever made the claim that misinformation didn't exist before AI so it's not clear who you're arguing with. People are rightly concerned about the scale of misinformation that AI is unlocking.
Re:Chatbot Hallucinations Are Poisoning Web Search by ryanklee on Thursday October 05, 2023 @08:11PM (#37785823)
I didn't say that they did.

What I'm responding to is the strong tendency to discount our very long history of dealing with factually incorrect information and the ascertainment of truth from sources both dubious and trustworthy.

Entire institutions are set up in order to handle these very real problems, the set of which currently dwarfs the problem of hallucinations in GPT.

From a social perspective, non-GPT falsehoods are even more insidious, because we are inclined to trust and believe those whom we like and are like us.

Again, people are in the habit of discounting just how much we are wrong in our everyday lives. The hallucinations therefore appear more singular than they actually are.

Re:Chatbot Hallucinations Are Poisoning Web Search by carlosjobim on Thursday October 05, 2023 @05:05PM (#37784296)
The shadow libraries.
Re:Chatbot Hallucinations Are Poisoning Web Search by robinsonb5 on Thursday October 05, 2023 @05:30PM (#37784526)
I suspect in the coming years the Wayback Machine at archive.org will become ever more important - always assuming it's not lost as collateral damage in their copyright battles. Indexing that dataset and making it searchable would massively increase its value.

My inner conspiracy theorist can't help wonder if the continued reduction in search usefulness isn't part of an ongoing deliberate disempowerment of everyday people - but my rational side says it's merely an unfortunate emergent behaviour of the systems we've built.

Re:Chatbot Hallucinations Are Poisoning Web Search by beebeepka on Thursday October 05, 2023 @07:47PM (#37785661)
Generative Wayback Machine
Re:Chatbot Hallucinations Are Poisoning Web Search by StableAlkyne on Thursday October 05, 2023 @08:18PM (#37785868)
Honestly, I would love to see a LLM based plugin that would take a website, remove all the tracking garbage and filler, and just give me back a no-frills static html+CSS site that looks like it was made in 1995.

"Oh, you want a guide to writing your own loss function for Tensorflow? Here's an FAQ that could have existed on comp.lang.python3.tensorflow"

Re:Chatbot Hallucinations Are Poisoning Web Search by creata on Thursday October 05, 2023 @09:53PM (#37786385)
So you want a... reader mode plugin? How would it use an LLM beneficially?
Re:Chatbot Hallucinations Are Poisoning Web Search by zirgs on Friday October 06, 2023 @05:14AM (#37788599)
LLMs can write summaries of articles.
Re:Chatbot Hallucinations Are Poisoning Web Search by thewakalix on Saturday October 07, 2023 @12:09AM (#37798942)
Does reader mode eliminate "filler"?
Re:Chatbot Hallucinations Are Poisoning Web Search by Peritract on Friday October 06, 2023 @04:36AM (#37788380)
More unnecessary technology doesn't solve the problem of too much unnecessary technology.
Re:Chatbot Hallucinations Are Poisoning Web Search by civilitty on Thursday October 05, 2023 @08:12PM (#37785827)
That would be the best thing ever. Just ask GPT to hallucinate an entire page based solely off the URL when it doesn’t exist in the database.
Re:Chatbot Hallucinations Are Poisoning Web Search by __loam on Thursday October 05, 2023 @08:16PM (#37785860)
It's a consequence of running search as a for profit system. When you're optimizing for revenue, the system priorities are different than if you were optimizing for user experience or true knowledge. Arguably that means that search should be a public utility but then you have to be able to trust your government with your searches and privacy.
Re:Chatbot Hallucinations Are Poisoning Web Search by thih9 on Thursday October 05, 2023 @06:32PM (#37785096)
I wonder how we'd test for AI contamination. And would there be attempts to sell a larger data set, one that pretends to be human generated, but instead is padded with some AI content.

Does this mean we'd end up with a finite set of verified human only data?

Would people start going through all kinds of offline archives via AI-gapped means, trying to uncover and document new sources of human input?

Re:Chatbot Hallucinations Are Poisoning Web Search by Agree2468 on Thursday October 05, 2023 @06:40PM (#37785156)
Right now is best time to buy encyclopedias.
Re:Chatbot Hallucinations Are Poisoning Web Search by visarga on Friday October 06, 2023 @01:03AM (#37787263)
Why, you can dl any version of wikipedia from before 2023
Re:Chatbot Hallucinations Are Poisoning Web Search by dotnet00 on Thursday October 05, 2023 @06:40PM (#37785157)
In some ways it already is that way. If I come across an artist I suspect is passing off AI generated stuff as their own (without using the tagging features the site has to indicate as much), an easy test is to just check if they've been posting since before 2020. If they have, and the style has recognizable similarities, it's clear that it's honestly human made.
Re:Chatbot Hallucinations Are Poisoning Web Search by dongping on Thursday October 05, 2023 @07:04PM (#37785339)
Actually, that's great in my opinion.

Assuming that semi-convincing misinformation spreads everywhere, people will finally have to find the original source of a certain statement, verify their "knowledge supply chain", and maybe use logic to evaluate every single statement made.

Re:Chatbot Hallucinations Are Poisoning Web Search by redstonefreedom on Thursday October 05, 2023 @07:28PM (#37785524)
I've said as much as to one extra incentive (besides retrain cost) as to why openai has frozen the training period for post-2022. I think it's trying to generate as much data as possible before itself has contaminated its training set. They'll effectively have a monopoly over it; it's interestingly a rare example of "we can only do this once, then it's forever degraded". You really want a clean & discrete start.
Re:Chatbot Hallucinations Are Poisoning Web Search by downboots on Thursday October 05, 2023 @08:40PM (#37785996)
Re:Chatbot Hallucinations Are Poisoning Web Search by bpiche on Friday October 06, 2023 @12:15PM (#37792704)
Hey, you're internet famous now. Check out the end of this boingboing post.

https://boingboing.net/2023/10/05/ai-search-chatbots-output-...