The “AI” label also indicates the solution is way over complicated and simpler ways to improve the product have been ignored. For instance, Confluence now has an “AI” chatbot. Search is still substantially worse than grep.
There's still a search for reviews, but I can't find one for Q&A or for both together.
What's the point of working your developers to the bone (as Amazon is famous for doing) if the result is going to be this bad?
I'd really rather have time invested in making the memory footprint smaller again.
Which is an incredibly misguided idea specifically in the context of a developer IDE... I mean I do understand the rationale and value for the job in other contexts. Not so much in this one though.
Not profits, just stock price. The evidence that anybody who isn't "selling shovels" has made any significant money yet with AI remains thin.
I can't remember if it was Google or Facebook, but in the last earnings season one of them claimed directly that AI has improved their advertising revenue, but the improvement was not terribly out of line from what these companies have been doing for a while (not like suddenly they popped out a revenue triple, it was not obviously out of line with their normal reporting), and I wouldn't be surprised they were doing exactly what I said in my first paragraph. I did not find their claim terribly compelling, even if it was completely true and not just a sop to the stock market. Nobody else is even claiming that AI has increased profits.
I'm sure there's some startups with some profits, maybe even very exciting amounts of profit for a startup, but nothing that would move the needle for established companies. In fact we seem to already be in the consolidation phase for the industry, the startups are starting to go under. If we use that as a timer for where we are in the hype cycle, this is underperforming compared to other bubbles; at this point in the dot com crash, while the sector was heavily overinvested it was also quite obvious to a calm observer not panicking about their portfolio that there was definitely a there there, it just wasn't ready to sustain that much investment. There is so far a lot less there there with AI.
This hype is going to prove disastrous for the entire technology, I fear. If it was treated as a more normal technology, it would improve, it would be experimented with, companies would learn how to use it, failures would be shaken out and successes doubled down on, and in 3-10 years it would be a healthy industry making reliable, good money with a bright future for growth. But the way the stock market went absolutely ape shit, now the bar is, in roughly 6-12 months if AI has not quintupled the profit margin of all of the already-largest companies on Earth, it is a failure, and so are those companies. We've got a lot of people now with all the incentives in the world to blatently lie about their progress because as soon as they are truthful they become personally bankrupt as their stock options tank. Not because of any particular aspect of AI, except that it wasn't a miracle "drop it into any process instantly see quintupled profits", and I'm not inclined to be too annoyed at any technology for failing to be that. And it could take AI down with it. Again. I'd expect the resulting "AI winter" to be relatively short this time because the technology really is becoming promising, but the medium-term prospects for the tech are probably a lot dimmer than they theoretically should be because of this disproportionate frenzy.
[1] https://thomasvilhena.com/2024/06/easy-wins-for-generative-a...
Very rare example though. As a matter of fact, I couldn't name another one right now.