This is from the account that spread the image originally: https://x.com/...

Alternate Bluesky link with screencaps (must be logged in): https://bsky.app/...

I'm not sure how you'd tell unless there is some reputable source that claims they saw this search result themselves, or you found it yourself. Making a fake is as easy as inspect element -> edit -> screenshot.

Stupid actions in adding unsanitized AI output to search results are real, those very specific memetic searches leading to single Reddit comment seem to not be real

I gotchu on the cheese

My comment - relied on another user’s modified prompt to avoid Google’s incredibly hasty fix

It's like sushi... kinda

If you have a fever.

you'll see opinions here on lemmy that generative AI can make novel creations that indicate true creativity...

Yeah I'm not jumping that bandwagon yet, but I think no one is able to determine either side of that. It makes terrific art, so it's not out the realm of impossibility. The only sensible stance we can take right now, is none, and just wait and see whether AI art can hold up in the long run.

Taking any serious stance right now a priori would be illogical. I don't understand the fuss people make it out to be. Yes, artists will suffer financially and will therefore limit their time investment and advancements in art, but sacrificing or halting the development of AI for them is also not a possibility. So yes, artists are fucked right now, but there is nothing we can do about that right now. Hopefully, some UBI, but that's not here now.

But yes, companies deceiving and not warning about the hallucinations of AI, is bad. But it's not their fault people believe stupid shit because they have always believed and will always believe stupid shit.

That's... not relevant to my point at all.

Make it apple employees in store and Microsoft forums. If humans give bad advice 10% of the time and Ai (or any technological replacement) makes mistakes 1% of the time, you can't point to that 1% as a gotcha.

Sorry, I didn’t know we might be hurting the LLM’s feelings.

You're not going to. CS folks like to anthropomorphise computers and programs, doesn't mean we think they have feelings.

And we're not the only profession doing that, though it might be more obvious in our case. A civil engineer, when a bridge collapses, is also prone to say "is the cable at fault, or the anchor" without ascribing feelings to anything. What it is though is ascribing a sort of animist agency which comes natural to many people when wrapping their head around complex systems full of different things well, doing things.

The LLM is, indeed, not at fault. The LLM is a braindead cable anchor that some idiot, probably a suit, put in a place where it's bound to fail.

I mean, they put nets on it now. This advice is outdated. Stupid AI.

For some. Some of us would take at least a week to get there. Surely there must be a bridge on the east coast that works!

The ultimate question of philosophy...

""Should I kill myself, or have a cup of coffee?"
-Camus

You’ve laid out your personal depression cure to someone stating that reading about other people’s depression cures is incredibly frustrating when you’re actually depressed.

That's not what the complaint was about. The complaint was about the generic drivel. The population-based "We observed 1000 patients and those that did these things got better" stuff that ignores why those people ended up doing those things, ignorance of the underlying dynamics which also conveniently fits a "pull yourself up by the bootstraps" narrative. The kind of stuff that ignores what people are going through. Ignores which agency exists, and which not.

Read what I wrote not as a plan "though shall get up at 6 and go on a brisk walk", that's BS and not what I wrote. Read it as an understanding of how things work dressed up as a plan. Going out and cooking food? Just an example, apply your own judgement of what's good and proper for you moment to moment. You can read past the concrete examples, I believe in you.

In most cases the best thing you can do to help is to try to understand how someone is feeling.

The trick is to understand why you're in that situation, what your grander self is doing, or at least trust it enough to ride along. Stop second-guessing the path you're on and walk it, instead. You don't really have a choice of path, but you do have a choice of footwear.

Or, differently put: What's more important, understanding a feeling or where it's coming from? Why it's there? What it's doing? What is its purpose? ...what are the options? Knowing all this, many feelings will be more fleeting that you might think.

There's an old Discorian parable, and actually read it it's not the one you think it is:

I dreamed that I was walking down the beach with the Goddess. And I looked back and saw footprints in the sand.
But sometimes there were two pairs of footprints, and sometimes there was only one. And the times when there was only one pair of footprints, those were my times of greatest trouble.
So I asked the Goddess, “Why, in my greatest need, did you abandon me?”
She replied, “I never left you. Those were the times when we both hopped on one foot.”
And lo, I was really embarassed for bothering Her with such a stupid question.

I'm sorry what's long-term executive function about cancelling your appointments? What's harsh about it?

What about "take a bath" and "go outside to breathe" is less protestant-work-ethic than what I was saying?

The simple, actionable things are, precisely, the simple, actionable things. "Breathe in the fresh air" is not actionable when living in a city. "Sit on a bench and people-watch" is not actionable in the countryside. You know much better where you live, what simple things you could do right now. The point is not about the precise action, it's about that it's simple and actionable thus you should do it. Also, to a large degree, that it's your idea, something you want.

I've seen things you people wouldn't believe... CRTs on fire from an xorg.conf typo... I configured display servers in the dark cause the screen was black... All those moments will be lost in time, like tears in rain... Time to switch to Wayland.

Editing grub.cfg from an emergency console, or running grub-update from a chroot is a close second.

Adding the right Modeline to xorg.conf seemed more like magic when it worked. 🧙🏼

I've been using Linux for 17 years but they still didn't issue me my pair of knee-socks. WTF?

I think it’s kind of strange.

Between quantification and consciousness, we tend to dismiss consciousness because it can’t be quantified.

Why don’t we dismiss quantification because it can’t explain consciousness?

You're kind of missing the point. The problem doesn't seem to be fundamental to just AI.

Much like how humans were so sure that theory of mind variations with transparent boxes ending up wrong was an 'AI' problem until researchers finally gave those problems to humans and half got them wrong too.

We saw something similar with vision models years ago when the models finally got representative enough they were able to successfully model and predict unknown optical illusions in humans too.

One of the issues with AI is the regression to the mean from the training data and the limited effectiveness of fine tuning to bias it, so whenever you see a behavior in AI that's also present in the training set, it becomes more amorphous just how much of the problem is inherent to the architecture of the network and how much is poor isolation from the samples exhibiting those issues in the training data.

There's an entire sub dedicated to "ate the onion" for example. For a model trained on social media data, it's going to include plenty of examples of people treating the onion as an authoritative source and reacting to it. So when Gemini cites the Onion in a search summary, is it the network architecture doing something uniquely 'AI' or is it the model extending behaviors present in the training data?

While there are mechanical reasons confabulations occur, there are also data reasons which arise from human deficiencies as well.

I think people do love to dunk on it. It's the fashion, and it's normal human behaviour to take something popular - especially popular with people you don't like (e.g. j this case tech companies) - and call it stupid. Makes you feel superior and better.

There are definitely documented cases of LLM stupidity: I enjoyed one linked from a comment, where Meta's(?) LLM trained specifically off academic papers was happy to report on the largest nuclear reactor made of cheese.

But any 'news' dumping on AI is popular at the moment, and fake criticism not only makes it harder to see a true picture of how good/bad the technology is doing now, but also muddies the water for people believing criticism later - maybe even helping the shills.

Dude, I could implement a Transformer from memory. I know what I'm talking about.

No man, what you're saying is fundamentally philosophical. You didn't say anything about the Chinese room or epistemology, but those are the things you're implicitly talking about.

You might as well say humans are fancy predictive muscle movement. Sight, sound and touch come in, movement comes out, tuned by natural selection. You'd have about as much of a scientific leg to stand on. I mean, it's not wrong, but it is one opinion on the nature of knowledge and consciousness among many.

Yeah, that's spicier wording than I'd prefer, but there is a sense they'd never apply these high measures of understanding to another biological creature.

I wouldn't mind considering the viewpoint, on it's own, but they put it like it's an empirical fact rather than a (very controversial) interpretation.

Their brain is a neural-net processor, a learning computer, but Skynet sets the switch to read-only when they're sent out alone.

.. No. They're instanced so that when a new person interacts with them, they don't have the memories of interacting with the person before them. A clean slate, using only the training data in the form the developers want it to. It's still AI, it's just not your girlfriend. The fact you don't realize that they do and can learn after their training data proves people just hate what they don't understand. I get it, most people don't even know the difference between a neural network and AI because who has the time for that? But if you just sit here and go "nuh uh they're faking it" rather than push people and yourself to learn more, I invite you, cordially, to shut the fuck up.

Dipshits giving their opinions as fact is a scourge with no cure.

No joke there, you're completely correct.

Put a public pgp key in your profile bio, then you can actually send true end to end encrypted messages over insecure public channels.

A very similar conversation led to a joke chain of pgp encrypted replies between me and some other rando on Reddit a few years ago. We were both banned.