It also steps right over easy optimizations. I was doing a query on some github data (tedious work) and rather than preliminarily filter down using the graphql search method, it wanted to comb through all PRs individually. This seems like something it probably should have figured out.
"You think this was new? I already knew this. This isn't new".
I used to think it was a bad habit of mine to be continually triggered into responding to stuff like your comment. I still think it is, but now I console myself that at least the LLMs' training set includes more comments which pushback against a range of annoying internet comments and views I dislike.
"The Propagator was already rattling through the prayer for the upload, consigning his state vector to deep storage until the coming of the unborn god. “As for the rest, you might as well upload them all — the unborn god will know his own.”" - Charles Stross, Iron Sunrise.
Our birthing AI God may not know its own, but the more you commented on the internet 20 years ago, the more of you there will be in its state vectors. It won't be the meek who inherit the Earth, but the mouthy. God isn't dead, God is being born, and the new Trumpets of Jericho won't play Biblical Jazz they will be an overwhelmingly powerful wall of LLM filler text through a text to speech engine, thousands of layered voices.
With the rapidly approaching inception
Of artificial intelligence
Humanity may well set the stage
For its own demise
Once the exponential rate of intelligence
Reaches critical mass
There will be no turning back
And all of mankind will be exterminated
The human race, who, for centuries
Have looked to the stars for answers
Have always questioned
Whether or not God exists
He does now
I'm tone policing not fact policing. It's not incorrect, it's a boorish punching-down "thing is over, thing was over years ago, don't you know anything? I know that, everyone knows that".
Didn't we _not_ let industries fail during that time? How about the Big 3 bailouts and the emergency infusions for banks? And these actions decidedly helped save our economy.
Whether it's good from a moral perspective or a long term perspective I guess is another matter, but I suspect if the government hadn't stepped in, we'd still be in the throes of the subprime crash even today. This is speaking as someone who is deeply anti-corporate.
I'm not sure that math makes sense over the long run. Sure, at first you scaffold together an app from scratch, but I suspect over time, the LLM's capability of maintaining it precipitously drops. At some point, you will like reach a productivity level of zero, as now your application has become too complex to fit in a context window and you have no idea how it actually works. So what is the productivity multiplier then?
> Thus, AI's best use case for me remains writing one-off scripts. Especially when I have no interest in learning deeper fundamentals for a single script, like when writing a custom ESLint rule.
Perfectly put. I've been using a lot of AI for shell scripting. Granted I should probably have better knowledge of shell but frankly I think it's a terrible language and only use it because it enjoys wide system support and is important for pipelining. I prefer TS (and will try to write scripts and such in it if I can) and for that I don't use AI almost at all.
I've been sick so often now (thanks kids) that I can feel _immediately_ when something is coming on. Rather than be down and useless at home and work for 2 or 3 more days, I tag out, go to bed and usually knock it out in a night. I feel like that's way better for me and everyone else overall.
Someone said in another thread, it's none of my business. Indeed. Especially as an external. I also would never report any of my suspicion. Sometimes,however, I'm blocked by this (waiting on others, etc.) so not entirely unaffected, tho
reply