My employer hired a consultancy company to implement some terraform and ansible code for some cloud infra and we got in return hundreds of machine generated code lines that do not work and now are suing the consultant.

Worst part, is that the guy who reviewed this ended up writing everything himself

I use AI as a rubber duck or to get general ideas all the time. For example, I wanted to make a “hand” of cards splayed out using css, so I asked AI and it gave me a nice starting point I could tweak without having to fuss with figuring the formula to tilt which element at which angle. It’s also quite good at guessing what boring boilerplate code I need to type next.

Another example: I was trying to figure out an architecture that adhered to OOP/SOLID principles for a specific task, and asked for an example implementation. I was able to test and think through a ton of permutations before landing on what I was taking to prod.

I think it’s a nice tool for the toolbelt, but it isn’t replacing a programmer anytime soon. You have to know what to ask and be able to intelligently analyze what it spits out to you

AI is good for the messy inbox problem, but it ain’t perfect.

Later, they comment:

I’ve learned today that you are sensitive to ensuring human readability over any concerns in regard to AI consumption

Their takeaway from

1. They reject with one of the reasons, and the only disclosed reason, being AI is worse at reading the form
2. Community says AI readability should not be a priority over human readability
3. That it may not even be a problem for AI to read
4. Suggest at least considering an improvement in a form that both can read well

is that the community wants to “ensure human readability over any concerns in regard to AI”.

I don’t think this is only about MS or being overworked. Yes, it was a harsh push-back. But they’re responding passive aggressively, claiming the community pushes the other/an extreme when, to me very clearly, it does not.

Maybe you can say that conclusion is also due do being overworked and not investing the time to read through the comments. But I dunno. There’s no need to reply in that passive aggressive tone and claiming unreasonable things.

If it can’t understand human text then is it really worth using? Like isn’t that the minimum standard here, to get context and understand from text?

I don’t have any game in this but this seems backwards and stupid… Especially since all AI currently is fancy pattern matching basically

Having AI not bullshiting will require an entirely different set of algorithms than LLM, or ML in general. ML by design aproximates answers, and you don’t use it for anything that’s deterministic and has a correct answer. So, in that rwgard, we’re basically at square 0.

You can keep on slapping a bunch of checks on top of random text prediction it gives you, but if you have a way of checking if something is really true for every case imaginable, then you can probably just use that to instead generate the reply, and it can’t be something that’s also ML/random.

He admitted himself that it might not be harder for AI to understand.