Is AI has a decent impact but is over-hyped?

It has an impact in multiple fields, but so far its not what the pro-AI absolutists thought it would be and I think its questionable even AGI or ASI will cause anything like the singularity. What do you think about AI?

lots of use cases and it's impressive so far personally. i use it for scripting (powershell, js, python) and summarizing information/books, health questions. its clear it's not going away and with time it will be more ingrained into daily life (see windows copilot). ultimately you need to be trained as 'prompt engineer' to take full advantage. otherwise you are like a baby looking at complex mechanism, thinking wow this thing can tell me the weather.

Over hyped, everyone is trying to do everything with AI even there is no need to use AI and even people don't have a real use have a hype for something they do not use at all.

MoonStarz said:

Over hyped, everyone is trying to do everything with AI even there is no need to use AI and even people don't have a real use have a hype for something they do not use at all.

Click to expand...

I had the same experience.

Like crypto, AI also over-hyped

Overhyped is when everybody want to jump on the thread and end up abusing the usage. The is common amidst the Youtubers, AI as given many people the privilege to own a channel now.

Of course it is over-hyped. People here are the ones promoting it! It's what we do.

That it can't do half the things people claim doesn't mean the capabilities it truly has aren't great.

And here we're almost exclusively talking about generative AI, which is a tiny slice of the overall ML and AI field.

I think the singularity is entirely possible. Even probable under certain circumstances. Whether it is desirable or not... I have a feeling we'll eventually find out if we live long enough.

djsobuj said:

Like crypto, AI also over-hyped

Click to expand...

Over-hyped like crypto, but it is not like crypto. AI has lots of uses. Crypto has... a few.

i've replaced an army of content writers with a few ai software and a few editors. cant imagine what its going to be capable of in the future.

AI is not overhyped.
It's the game changer, but only if you don't think about AI in ChatGPT category only.

Graphics production, video production, custom language models, programming - here AI do it's magic.
It can be overhyped by tech bloggers, who don't know what they are writing about and everything is "ChatGPT".
In the meantime, in developer basements, people are working on revolution.

It actually has a very serious effect. A lot of people have lost their jobs and it is going to get a lot worse soon...

wimiphp said:

It has an impact in multiple fields, but so far its not what the pro-AI absolutists thought it would be and I think its questionable even AGI or ASI will cause anything like the singularity. What do you think about AI?

Click to expand...

AI!=LLMs. But currently only transformer based LLMs. My opinion: right now overhyped. AGI and singularity far away. But heavy impact in next coming years - rather in 2 than in 5 years. E.g. being in university, studying IT and becoming soon junior software developer? Uuuh, good luck ) There is a lot of academic and fact-based discussion ongoing, describing the limits of transformer based technology for LLMs, the need for more data and intellectual scalability and showing that the current approach inherently barely can create a new Albert Einstein. But he is not your competitor in your job.

Lukmat said:

AI is not overhyped.
It's the game changer, but only if you don't think about AI in ChatGPT category only.

Graphics production, video production, custom language models, programming - here AI do it's magic.
It can be overhyped by tech bloggers, who don't know what they are writing about and everything is "ChatGPT".
In the meantime, in developer basements, people are working on revolution.

Click to expand...

I think there are enough bloggers and other influencers who know that AI does not include only ChatGPT, but even than all the niches you mentioned still do what ChatGTP does, using mainly transformers.

In regards to the revoltution, the only likely revolution is going to be the one that has already taken effect and still somwha going on, which is the inooperation of the said transformers and even then their use case is more limited than people want to admit.

Form what I observe most people that see a revolution due to AI are not the developers but CEOs of the big tech companies, who also don't all agree that there will be a revolution given Zuckerbergs video as an example, in which he talks about making even AGI open source. This could be intepereted in several ways, inlcuding that either OpenAI and other AI companies likely overhype AI to get more sales or/and even AGI will just be another tool we can use for he already existing compuer/internet infastructured.

In other words, there seems to be a lot of hype, but even that hype seems to have mainly cooled down to the general public at least, because most AI companies have basically only changed the modalities and upgraded their nlp models slightly to stay relevant, but I think we likely see more integration of the exisitng AI and possibly AGI over the next years if not decades than anything else and more people will realize that not everything profits form AI due to its limited scope and espeically the models having huge expenses and the open soure comunnity can either not keep up, or will get lawsuits all over the place for reuising big companies architectures. Not even to mention several other problems, like using huge AI systems in companies that value their privacy, which most mid to big tech companies do if not including the smaller ones, which likely also value their data.