Is the AI Boom Already Over?
vox.comWe've kind of been through this already with autonomous vehicles a decade ago.
In a demonstration capacity autonomous vehicles were mindblowing. In the real world the novelty wears off quick, and what remains is a bit of a disappointment.
Of course, the demos helped churn up many bajillions of dollars in investment money, and many companies with long runways continue to tinker away at the problem. Level 4 AVs are dramatically better than a decade ago, and yet still totally shitty relative to real to real drivers.
What has become apparent is that these companies and their investors are not at all interested in solving any real transportation problems.
They are interested in getting between you and things you want to do, so they can set up a gate and charge you money to pass through it.
> the bots are still prone to basic errors that make them impossible to trust
This has been my frustration. If I don't bring my own domain knowledge, carefully review, and provide follow up prompts I get garbage output. Novices don't have great domain knowledge and likely overestimate the quality of the output. It's not worth the effort.
It's helpful to know what chatgpt can't handle: math, visual, spacial, safety. It's pretty limited.
I'm trying to remain impressed and open-minded, and there's no question these things are super cool, and something genuinely exciting. And who knows how much they'll improve over the next few years? But so far, if I'm brutally honest, they have largely been frustrating to work with. I am a teacher, and I use them (mainly via APIs) to generate questions for worksheets and study guides for my students. The idea of letting the computer help me save time on creating those kinds of exercises so that I can then focus on the more interesting/challenging activities initially sounded so good, but I've found I'm completely unable to trust these systems to generate correct information. Almost every question set or study guide I've generated had at least one, sometimes several, serious errors that would have led my students astray. It does still save time, and I'm sure it will get better. But like you said, the fact that I have to constantly correct its output basically eliminates the possibility, at least for now, of using it as a tutor to teach me material that is genuinely new to me (something I was looking forward to doing). I simply can't trust it for that at this point. We'll see how things evolve.
Sure, you can define "the AI Boom" in such as a way as to say "The AI Boom is over." No problem. But... taking a step back and looking at the bigger picture, I'd say that any slight, contemporary, dip in interest in AI is just that: a dip. A momentary blip at a point in time, hardly the harbinger of any larger trend. There's no reason I see to expect anything less than for AI research to continue, and for AI to become progressively more and more useful.
So no, not in an bigger, more universal sense, would I agree that "the AI Boom is over". If anything, it's just starting.
It can't be over because it never arrived in any real, substantial way.
We've been waiting for some real "intelligence" from AI for decades now.
Creating word salad from ingredients found on the internet isn't it.
I work in the field and I don't want to say this is completely correct, but there are large swaths of the technology and hype for which it's basically true. There are lots of cool tricks and lots of promise but, in particular with generative AI that isn't a generic chat assistant or text to image model, there has not been a demonstrated business case with any traction (I'd invite anyone to give me a concrete example that disproves my assertion). I think there are promising avenues in the tech and I'm working to change this, as are many others. Though I'd say most of the people that are heavily promoting it are trying to sell some generic platform or tool or something that assumes the buyer has their own use case for it.
Anyway, ML has some uses and the latest round will too but most of the hype will be unrealized, as it has been with computer vision, deep learning, etc.
Ha, no I don’t think so.
This is how we experience the slope of technical change over a period of months: ___
This is closer to the slope of technical change over that time period:
/