It's Time to Believe the AI Hype | Hacker News Reader

It's Time to Believe the AI Hype

wired.com

8 pointselsewhen9 comments

Not a single word about user privacy, which is more than a little ridiculous.

Even if you agree with the author, it is impossible that all of these impressive capabilities that depend on recording everything about your life will exist without sending the data to the cloud.

I don't care how good your brand is; I will never trust any company with the sort of information they'd need to deliver an AI assistant like this.

Open source / local AI is now about as capable as GPT 3.5 which is about a year and a half old.

I think we may see an open source gpt4 equivalent in the next 18 months. Anthropic has promised to open source Claude 3, for example.

If that happens, we could get good, fully private open source AI .

That'd be nice. Local models are absolutely improving but I believe there's going to be a capability plateau for generalized language models regardless of where they're hosted.

GPT4 is good enough to have profound effects on my quality of life as an autistic contractor/founder struggling with self advocacy and executive function disorder, and homelessness.

It's kept me off the street, and out of burnout for over a year, and has let me implement a lot of improvements to my business process.

People who think GPT needs to get any better to have impact simply have not spent enough time working through serious problems with it, and have not read prompting papers.

If I can get a free and private GPT4 replacement, that wil be one of the most life changing events of my adult life. I've struggled through two decades of homelessness without the kind of emotional and executive support that GPT now provides. It helped me reconnect with estranged family and loved ones. It helped me rebuild a social life. It helped me find housing. When I could not decide what to eat or when I did not have enough food it helped me take action and feed myself. I feel a new hope in my life, and a new and effective way to overcome many of my problems.

So I'm going to fight very hard to make AI a fundamental human right!

I've let myself get behind on AI; so, I'm not sure how it could help with all that. I'd be interested if you could point me to some articles and/or explain. Especially, the "emotional and executive support". I have ADD and struggle with lots of things and so, I understand executive brain function; I assume that's what your referring to, but correct me if I'm wrong.

"working through serious problems" I would love to know how to do this. Is it just asking it a question and/or a series of questions?

I tend not to do subscriptions besides phone/internet (the same thing, really, I believe)

I may do it today. Regardless, I understand chatgpt is integrated into bing, but it doesn't have "plus" features. I may have to read more, but maybe I'll start with bing. Anyway, let me know please! :)

Thanks

That article is breathless and uncritical. One thing I took away from it is that the author certainly believes the hype.

https://archive.ph/11fWX

Reply on news.ycombinator.com