This exactly. LLMs can't reason, so we shouldn't expect them to try. They can do translation extremely well, so things like converting descriptions to 90-95% correct code in 10-100x less time, or converting from one language to another, are the killer use cases IMO.
But expecting them to solve difficult unsolved problems is a fundamental misunderstanding of what they are under the hood.
I picked this problem specifically because it's about "converting from one language to another". The problem is already solved in the literature. I understand that doing cutting edge research is a different problem, and that is explicitly not what I'm doing here, nor what I am expecting of the tool. I have coauthored an actual published computer science paper, and this excercise is VERY far from the complexity of that.
Could you share some concrete experience of a problem where aider, or a tool like it, helped you? What was your workflow, and how was the experience?
It's not about avoiding hard work - the audience on HN skews wealthy due to heavy representation of skilled devs in their 30s+, but the average person does not earn anything close to FAANG salaries. Even most devs in general don't earn like that. The interview process being fairly well understood in general, any advantage that can possibly get a person from $60k/year to generationally-life-changing $300k/year will be used eventually.
Disclosure is technically required, but in practice I see undisclosed ads on social media all the time. If the individual instance is small enough and dissipates into the ether fast enough, there is virtually no risk of enforcement.
Similarly, the black box AI models guarantee the owners can just shrug and say it's not their fault if the model suggests Wonderbread(r) for making toast 3.2% more frequently than other breads.
It also extends to IRL, there are mansions you can rent by the hour to film content as though you live there. That's a big reason why so many influencers are in samey giant empty kitchens for their "stupid food" videos where they make a giant mess - it's not their home.
Ditto the above for luxury cars, watches, anything that can be rented for show frequently is, especially when ostentatious displays of wealth are core to the message, like finance/crypto influencers peddling courses or coins.
Yikes I hope this is tongue-in-cheek, I definitely don't want a statistical process deciding whether to surface a life-critical alarm to healthcare staff
It's the same as allowing full self driving cars which on average are safer than human drivers but sometimes accidentally drive into a fire truck because they couldn't train an image classifier to more accuracy than 99%.
"I'll tell you one thing about the universe, though. The universe is a pretty big place. It's bigger than anything anyone has ever dreamed of before. So if it's just us... seems like an awful waste of space. Right?"
The most telling part about this "debate" is conservative activists blasting center and center-left outlets for barely off-center comments, while ignoring the blatant extreme partisanship coming out of right wing media outlets like Fox.
Why do these "activists" always seem like they're operating in bad faith?
Pretty sure it's because this demo floor is so small - in the clips of two people on the mat at once you can see if they took larger steps they'd physically step off the mat. I hesitate to assume what it feels like, but I imagine larger area mats may allow more natural steps
I used to work on call in non-software, as a blood analyzer Field Service Engineer for medtech companies, like Siemens, Roche, Beckman-Coulter, Bio-Rad, etc. I spent 5 years doing this.
In that industry and job it's standard to work on call rotations that last for 7 days through the weekend and repeat every 3-5 weeks depending on team size. Because the job is hands-on and at customer sites, the minimum response time is usually a bit more than 15 minutes.
If anyone is considering accepting an on-call rotation, don't. It takes over your life. At first you may think it's fun to swoop in and be the hero fixit person, or maybe you at least tolerate it because of high base pay or something. Eventually you grow to hate the sound that accompanies the callout text or email. By the end, it triggered a visceral physical negative reaction I later learned was a minor anxiety attack.
The thing managers seeking to add on-call don't understand (those who haven't done it themselves), is that when you're on call you can't do anything. Not only can you not drink (minor annoyance), you can't even leave the house for normal stuff like a movie (medium), or attend events like a kids graduation or recital (major).
And before you say "just swap rotations around on your team," I worked with about 10-15 guys and you'd be surprised how hard it is to swap. Everyone makes plans for things all the time without realizing it would interfere with on-call, like going to the gym. At the start of the year when the on-call rotations would be released to us, there was a mad rush to submit vacation requests to get it locked in before illness or other things required swapping on-call around, much less having to take field requests from colleagues that you may (will) have to rely on when you in turn need a swap, so you can't just ignore every request.
This manager and CTO are hilariously unprepared and uninformed, bordering on pure dreaming. A 15 minute max response time is criminally unrealistic, you can't even go to the grocery store. And once a week seems low on paper until you realize that depending on team size, you can easily wind up being forced to give up a quarter of your weekends for the year (as I had to).
I haven't worked on-call for software, so maybe it's different, but I doubt it's really too different.
After 5 years of all that lost time, the amount I'd have to be paid to go back on-call is so astronomically high no company would ever pay it. Actually, now that I have a kid, there's probably no amount I could be paid if it means missing time with them.
It depends though. I totally see your point and when I was working in a company where we spread the oncall between all people it was a) usually no problem to swap a few hours (we did full weeks, so if you had one thing for 4h one night, usually someone would cover) and b) our response times weren't quite SO harsh - so we did some things, as long as you had cell coverage and a laptop nearby. Of course the cinema was out, or being on the road for hours, but going for groceries was no problem.
I still hated it and don't want to do it anymore because we had so many factors out of our control that made "false" alarms where the only thing to do was check if it was really a problem and then open a ticket upstream at a vendor...
It seems like OP and their CTO would like to have reliable 24/7 operations but do not have the skill to do it by designing systems that won't fail catastrophically out of hours or the budget to do it by hiring dedicated staff to be available at all times to fix problems. So instead they are trying to abuse their existing staff.
The correct response to this IMHO is for the existing staff to collectively tell management where to go. In life you can't always get what you want if you don't have the money or make the effort it requires.
But expecting them to solve difficult unsolved problems is a fundamental misunderstanding of what they are under the hood.