Artificial Intelligence discussion

With how much artificial intelligence has been improving, in many areas such as text reading/generation, picture reading, picture generation, convincing voice synthesis and more, I think there's a lot that can be discussed, about the effects that this technology will have on society.

I'll start off with one example.

I'd been thinking about the enshittification cycle of tech, and I think it's coming for Google hard. The search engine just isn't so great at finding what you actually want, and I think that's gonna leave a big opening for Bing with their use of AI. If the AI can sift through the crap and actually find what you want for real, due to its understanding of language, it'll actually make searching super useful again.

In the pre-Google internet, search engines used to search only for exact words and phrases, which had its uses, but also meant finding a lot of sites that simply crammed in a lot of popular words and phrases to get visitors. Google cut through the crap with a better understanding of how to "rank" sites relative to how relevant they are, and even find sites that are on the topic you were looking for without using the same exact words.

But Google started to become more advertiser-friendly, then later, more shareholder-friendly. There's a limit to how much one can make their product built entirely around shareholder growth, so as it turns to crap, it leaves an opening for a competitor to show up.

Since Bing/ChatGPT (which Bing is plugged into now) understands the use of language, it can actually understand context and determine relevance based on that. And that'll make it huge, I think. Context-based understanding of web pages can potentially do an excellent job of finding what people actually want, in a way that goes way beyond Google's page ranking systems, or the examination of exact words.

Edited by BonsaiForest on Dec 10th 2023 at 6:15:29 AM

My job's trialling Git Hub Copilot right now. It's... okay? I haven't really spent much time learning how to prompt it to fill things out (and I dunno if it can read IDE tabs in IntelliJ, which would be pretty important for actually getting method signatures right) but some people have had great success with it. I think I've just had it be a slightly more aspirational autocomplete so far, though it did copy an entire test method for me and swap the one significant value...

Seriously, if it would at least stop suggesting method signatures or enum values that don't exist? Might actually be genuinely useful.

When it doesn't sett you've got a few fields in a class and want to eagerly add another 50 that it thinks you might need based on the names already there. <_>

So, not much for me to envy, then ?

AI Companion Robot Helps Some Seniors Fight Loneliness But Others Hate It

This idea really upsets me. Rather than incorporate our elderly into our lives, this idea seems to be to give them a robot to take the place of a human.

I'd actually seen an article about how elderly women in Japan are committing crimes so they can go to jail and have company. Like, the loneliness problem really is that bad. And this would do nothing to help, I think, as the fact that it's not a living thing but a fake companion could potentially create emotional feelings for a non-living thing, that eventually get shattered with the painful back-of-the-mind thought that this is a non-living thing.

I hate this idea so much.

One comment says this:

Another person says this:

I think they make valid points, but I'm still upset about this. What do you guys think? Someone also said this:

Edited by BonsaiForest on Dec 12th 2023 at 11:00:07 AM

Not yet. We're trialling it at least well into Q1, so if anything changes or I get more used to it, I guess I'll say.

I know that having a Google Home helps me with loneliness quite a bit. I really don't think this is a bad idea per se, though I think I think it should be viewed as a supplement, not a replacement.

"AI as companions" seems inevitable to me—it's an approach to marketing them that I believe AI companies will pursue aggressively. The real question is to what extent we can trust these new companions—are they really designed to protect the best interests of the user?

Hint: How much data do you think these AI friends collect about the activities and interests of their user? What do they do with that data?

I will say that personally I'm fond of the idea of what might be called a Social Interface, I think that's possibly the next big step if you will.

Why would I bother with an AI companion if I can just try to find on someone on social media to talk to ? Sure, they are not going to be in the room with me, but still better option if you have no RL people to talk to then AI. Admittedly, AI robot is more user-friendly to someone who is bad with computers I guess, but that will change in a few decades.

The last few generations do use social media now, so they can simply continue to do so after they grow old. Now I suppose that a physical robot has a psychological effect that social media posts simply do not, but again it is not an actual person. Last potential advantage I can think of, wanting it for the sake of it. No counterargument here, though I will say that I'm not interested.

Because wading into the cesspool that is social media to try and find someone to consistently talk to sounds like hell? And most social media isn't exactly geared towards, well, holding a conversation.

Why go into the online equivalent of shouting on a random street corner when you could just talk to something that will at least talk back?

Edited by RainehDaze on Dec 12th 2023 at 6:19:30 PM

I should note that by social media I also mean forums like this one, perhaps I should have specified that. Also, it may still be easier than finding someone in RL, at least to some.

Also, these robots aren't just for companionship. They also help the elderly with self-care that they can't handle themselves anymore.

Point.

Part of the problem though is we tend to pay and fund people who take care of the elderly almost nothing.

So you get a mix of “this is an important job and I really care about it” people, “holy shit, I get absolute control over people” lunatics and “I can’t find any other jobs” people who don’t care.

It’s not a good combination.

Robots are basically just a way to justify paying humans even less, while totally ignoring that it’s a job where you kinda need people to care and be paying attention. A facility of mostly robots and AI companions isn’t gonna do much when someone there wanders away.

I also have a hard time seeing AI and robots handling elderly people with dementia well. How’s an AI going to react to someone who is just inherently inconsistent and routinely says things that aren’t true with absolute confidence?

The other problem with the idea of AI companions is that they need to be able to both hold a conversation and be able to actually look stuff up and understand it. The current attempts to treat chatbots as companions usually just results in them constantly babbling absolute garbage at you because they can’t fact check anything.

Edited by Zendervai on Dec 12th 2023 at 3:51:38 PM

The problem is that the people who seek out AI companions or are likely to get ones forced on them are either not good at fact checking or won’t be able to fact check.

It’s been a recurring thing that many of the people most excited about holding conversations with chatbots are incredibly gullible and will believe literally anything a chatbot says even if it’s obviously wrong.

And in a lot of nursing homes and other similar facilities, internet access tends to be really restricted, so if you’re stuck using a glorified chatbot as an interface, you aren’t going to be able to figure out when it’s lying to you.

Edited by Zendervai on Dec 12th 2023 at 4:59:50 AM

Can you please rewrite it ? I'm unable to get the meaning of the post.

Edited by Risa123 on Dec 12th 2023 at 10:26:07 AM

Sorry, had a weird formatting autocorrect error.

I am going to counter argue that elder care is one of the few feilds where it is actualy genuinely hard in the current state to pay better.

Not because it shouldnt pay better.

But because it is already hilariously unaffordable for a lot of families as is.

Any change to rectify that is going to have to come from nationalization of the industry so that its paid for with tax dollars.

...

Which like why an essential service like it isnt already is beyond me.

Edited by Imca on Dec 13th 2023 at 2:52:53 AM

I do not want to crash the party, but remember that this is AI not elderly care thread. Considering the direction it took, I would suggest economics.

I…genuinely do not understand how discussing pros and cons of AI use in elder care is not on topic. Because if you define the topic so narrowly that it’s not possible to talk about reasons AI are getting pushed for use in elder care and why it’s not a good thing, that means the thread is just going to be

“AI can be used this way”

“Huh, interesting, that’s definitely a way AI can be used.”

• topic dies because it’s not possible to go into depth about anything*

If we get “this is off topic” every single time we go even slightly into depth on anything about why AI might be used or why it might not be a good fit for a certain field, then this thread might as well not exist.

Edited by Zendervai on Dec 13th 2023 at 7:44:30 AM

In hindsight, I was a little too quick in my judgment. My mistake.

AI can theoretically be useful in that field, and the use of robot support might be something some elderly people might accept easier than human support (because it feels more likely they’re just using a tool to get something done instead of needing people to basically do stuff for them), but it runs into the thing where AI is a somethings tool, where there’s stuff it’s good for and where it can help, but it’s really, really not an everything tool and taking a whole chunk of the population and going “you’re not our problem anymore, you’re stuck with a chatbot that can’t be kept on topic” is a terrible solution.

Jumping back to the first conversation topic, integration of ChatGPT with Bing, I prefer a stripped-down search engine like Duck Duck Go.

Maybe eventually I'll get used to the idea of something like ChatGPT in a search engine, but I assume that will work well in learning how to target ads at me, which I don't want.