Is anyone else scared of how fast AI is advancing?

Anytime I ask it about things I’m knowledgeable about, it falls on it’s face. It only seems smart on topics I don’t know shit about (this red flag has spotlights on it!). Given this, I cannot trust it on anything lest I be led down false paths. Given how LLMs work, I don’t see this problem as fixable.

I have found them useful for search term generation. (Using a topic I know things about) If I need a more powerful drill for screwing lag bolts into hardwood and I’m not finding what I need, asking it and and getting back the keyword ‘impact driver’ is helpful, as it lets me go search for what that is and if it is the solution to my need for a more powerful drill. Note: I do not let it teach me about impact drivers (as it falls on it’s face all the fucking time), only use it to get the keyword to then use to search the internet.

To circle back to your question, I’m not scared with how fast they are advancing. I’m scared by how many people think they are good at everything and put them in places they don’t belong.

Not really. If anything, I’m frustrated how slowly it’s improving. Most new models are just made by pouring more money into the same problem instead of actually inventing anything new.

We have this glorified autocomplete feature and we are treating it for some reason like it would be an artificial general intelligence only because it outputs human like sentences. At the same time, it weakens the quality of the school system, products and services, raises the prices of consumer electronics, fills my internet with slop and destroys natural resources. We are living in a some kind of transitional phase and there are still many problems that have not yet been solved and it’s really frustrating.

no, it’s a glorified calculator. it doesn’t think.

It is scary, less because of AI itself and more because of corporate greed and regular people without critical thinking skills that will blindly trust anything AI tells them. AI is and can be a really useful tool for a lot of scenarios. I use it as a coding assistant for my job almost daily. But I dont blindly accept what it tells me or programs. I test, debug, and research things. Its good for summarization, easy, repetitive tasks, and taking suggestions from the internet.

AI is a tool, and like any tool, it can be really dangerous if you don’t use it in a smart way.

Not really…

… but you know what else is advancing? Lemmy! The 1.0 beta is out :D

The feeling of constant acceleration is valid. It will only get more intense in the coming years, so consider this your practice round.

In spite of the marginal results a lot of current LLMs can deliver, they are making a difference in some areas of work. I’m a data engineer and working with an agent over the last few months has been a revelation. You have to wrangle it just so, but with an adequate context well-defined, you can plow through months of tedious due-diligence and fine-tuning in days.

I wouldn’t trust it for medical advice… yet… but the time will come when it stops being “ai” and becomes like autofocus or voice transcription or shopping cart suggestions, just another tool. Something else will take on that mantle, and be a different, even more disconcerting mixed bag.

There was a book that came out about 20 years ago by Ray Kurzweil, named The Singularity is Near that discussed this phenomenon in detail, and so far has been prophetic. It will help you understand what’s happening and what’s coming next.

You should stop reading marketing junk. Several of the newer models have benchmarked as worse then original releases. Many firms roll back AI deployments, some models create more errors then it would take to fire and intern and some can’t even run a vending machine business

I’m concerned about the resources it gobbles up but I’m not skynet scared.

The effectiveness and the benefits are overhyped by the people with a financial interest in it. We call it AI but it’s only one for two there in that it is artificial. It’s a technology in its infancy. And it hasn’t found a use case that will guarantee an ROI. And the companies involved here are running out of runway.

I genuinely believed that AI this capable wouldn’t be feasible in my lifetime.

If you knew how things really were, you would know since the 2000s or even 1990s that such AI is quite possible, if you heard from scientists in the field of AI, the truth, of course.