Oh the number of time LLM will, for example, be giving me the list of bugs it found in code, when I ask it for a review, just to decide there’s no big half way through explaining it.
I think the problem here is that LLMs aren’t really “intelligence models” but more like “knowledge models”. LLMs don’t “think”, they just use a clever trick to make it seem like they do.
I might not understand a lot about current state of AI, but that’s what they seem to be. Give it information and ask to organise it and make links, and they’ll do it, but that’s it, they don’t continually try to get out of the knowledge box they’re at, they don’t even know there’s a box.
Even getting Amazon tablet that’s specifically designed for children comes with useless apps you just cannot get rid off, because they need to shove all their services into your mouth hoping you will accidentally swallow at least some of them.
There used to be time when every company had a mission statement and some (sometimes silly) ambitions, today it’s all just money, nothing else, not even a pretending anymore.
Silicon Valley is working round the clock to make bad behaviour as cheap and easy as possible, while taking all the profits. PR floods are probably the most innocent of this. Imagine if you want to send someone abusive text, previously you’d have to spend time and energy writing insults, now the machine can text or even phone someone on your behalf and harass them all day all night. Being a pest to the society was never as easy as right now.
> So, essentially, my experience is that it's automated away the easy bits
Yes, which is great for “easy” but mundane and time consuming tasks like refactoring structs when you have a ton of tests that look at snapshots of nested structs etc.
Squeezing that last few bits of “make comments consistent with the style and tone of existing comments”, or “stop adding docs that make it sound like an academic paper” or just keep the overall feel of the code same across all files take a lot of effort and energy.
For me the biggest issue with AI coding model is not that it writes code, or how it writes code, or whether it will replace me or not. I mean, these are real problems, but just not the ones that makes the difference to me daily. Instead what sets my feeling about the AI is simple: the experience of working with AI models, because it's the worse experience of my life so far.
Every single LLM will make up stuff, go down rabbit holes I didn't even ask it to visit and infuriate me all day long by doing ALMOST what I asked it to do, but just not quite. Take "yes" to one question as an approval to some other question. Come up with a plan but one you review the plan and accept it it hits some minor issue and then throws away the plan and do whatever it wants. Or when I reject some line of code or don't approve a command it wants to run 9/10 times it just tries to re-add or re-run it as if my rejection was meaningless. And no amount of rules and markdown files ever seems to change its behaviour for long, if at all.
It's like working with a sociopath who just doesn't remember anything that happened 5 minutes ago.
It's the classic case of "my computer doesn't do what I want, only what I tell it".
And all of that would be fine if it didn't pretend to be a human, if all the UX didn't create the illusion that you interact with some intelligent being, because as long as I remember that this is a cli tool it's all good, I manage my expectations, but the experience sooner or later makes you annoyed and frustrated - and if this was a person they would either stop or you wouldn't never work with them again.
I am constantly confused at work that people see AI as binary: it’s either a person writing the code and taking ownership of it, or it’s AI and the poor person has no control. It’s rather bespoke or entirely vibe coded. Why? When I use AI it’s still me writing the code, I just get better autocomplete and a tool that can verify my changes using fuzzy sentence as the linting rules. Why did so many people just immediately assumed that it all must be vibe coded now? Maybe it’s because everyone hates reviewing code they didn’t write?
Partially because they think that once employees put enough context and rules into markdown files they can be fired and replaced by an online subscription.
IMHO it’s just good old fomo. They fear that the landscape will shift overnight and they’ll be left with a ton of useless meat bags instead of a pay as you go models that you can upgrade instantly every quarter.
The only thing that interested in is: did they fix screen brightness getting super dim even when the slider is at max? That’s incredibly annoying and frustrating, and it’s been like that since first 26 release.
And it’s a clear bug because brightness resets to expected level if I go to photos and open an HDR image.
I can wait for autocomplete that doesn’t suggest garbage 50% of the time, but this one is just too annoying.
One can ask, what is a practical difference between “Check everything that it produces” and “Do all the work yourself”?
It’s not typing that’s the bottleneck, at least not often, so this is essentially assuming that you can do all the needed work without actually doing it, which is obviously wishful thinking.
All companies are now AI companies.
Just like a while ago all companies were suddenly Ads companies.
The entire tech sector is one big FOMO - once you reach certain scale you do exactly the same thing as everyone else.
I have no idea how much I’ve spent, it’s invisible to me, the company doesn’t share it with me.
I have no idea what “1 credit” means in terms of $$$, is that 1$? 0.1? 0.01? Is it even a fixed price?
I have no idea how much will given take cost. Well, I can ask for a plan and extrapolate from that, but all perfectly reasonable looking plans eventually end up in a rabbit hole.
Providers keep introducing new models and each is more expensive while offering modest improvements, it’s a silent inflation.
So I personally can easily believe that. Especially that a lot of people will just try to see if model can make that huge improvement / refactoring they’ve been hoping to do a reality, or tons of experiments to validate ideas.
I have an ancient Withings scale and a Fitbit, it used to be that Withings sends data straight to Fitbit but these days seems gone, so now I have Withings app writing to Apple Health and then Google Health reads from it. Great time to be alive
Call me old but there was a time when “open source project” meant “I had a problem, this is my solution, if someone has the same problem then you are free to use my solution”.
These days is more:
- building personal brand
- showcasing your skills
- trying to outsmart somebody else, often because they didn’t merge your pr
- sometimes just having fun
And if you work for big org it’s also often “this looks vaguely similar to one of our epics so let’s start using it and demand 24/7 support”