- cross-posted to:
- [email protected]
- cross-posted to:
- [email protected]
Now I understand why at each windows 11 update, they introduce more bugs than ever
You must log in or register to comment.
They include tab complete of github copilot which is often as much as a single dot. Same thing they’ve done with all github copilot stats.
If they mean “30% of the code we wrote last month” then I might believe it. Though I bet it is not across the board but deep in one or two areas. Still, it’s a crazy number.
But he said something like “30% of the code in our repositories” which would mean everything, including their entire legacy of code. And that I simply do not believe.
My first thought on reading that is: yeah, like about 98% of the human genome is “junk DNA” that we have little or no idea what it might be doing. Sometimes when we cut it out, nobody ever notices, sometimes when we cut it out the system won’t boot up.
It wouldn’t surprise me at all if they entered the entire codebase for Windows 11 into an LLM and asked it to optimize it or some shit lol
Yeah that’s a good point.
lmao I just said the same thing before reading your comment
And surprise surprise, it’s worse than ever
Of course it’s just bad writing, but I kind of wouldn’t put it past management to try shoving their multitude of codebases through an LLM at this point.
Its a shit article with Tech crunch changing the words to get people in a flap about AI (for or against), the actual quote is
“I’d say maybe 20 percent, 30 percent of the code that is inside of our repos today and some of our projects are probably all written by software”
“Written by software” reasonably included machine refactored code, automatically generated boilerplate and things generated by AI assistants. Through that lens 20% doesnt seem crazy.
The A stands for Automation, right?
I’ve been “automatically writing code” for a system of about a dozen modules - we specify a glue file in .json between all the modules and the code generating software makes units to go in each module to do the communication interfacing based on the glue spec. That system has been running for more than 10 years now, it writes a couple hundred thousand lines of “new code” every time we modify the glue file.
It shows
That explains so much
Beat me to it.
Year of the Linux desktop
(Amusing sidenote: my autocomplete’s first suggestion after ‘Linux’ was ‘propaganda’.)
I think it might actually start coming to fruition now that a lot of games have native support.
One of the only things stopping me from taking the plunge at this point is laziness to soft through all my data and make sure what i need is backed up before firmatting (i know, i need a good backup solution; open to suggestions here)
That’s… not something to be proud of.
If they start with those products today with zero marketing budged and zero user base nobody would use it. Those CEOs are just clowns.
Even worse, whenever a good new technology does pop up they buy it and ruin it.
How much of Linux is?
If you count all of my contributions, 0%.
None of my contributions have been included. I am a terrible programmer.
50% of my code is written by Intellisense…
Yikes
Boy am I glad not having to touch their software.
this makes way more sense than hundreds of shitty devs.
this is why I get so much business as a IT consultant lol
Even their AI crashes all the time, its brutal.
It shows
Stole it as if I wrote it
Copilot. Piloting you towards effortless bugs, and with all the telemetry, we don’t need to test our patches and updates. You, the user are doing that for us. Sincerely, Microsoft.
That’s why their products are so crappy!
What? There products have long been shit ever before AI was even a thing.
Anyone remember windows ME? I sure as fuck do.
No way, they get their results through honest effort. Anyone can make crappy AI products now, but Microsoft have been doing crappy products the hard way for decades. Don’t downplay their hard work !
I bet they’re counting code written while someone had an AI plugin installed as “written by AI” and I bet that accounts for almost all of that 30%. On top of that, I’m betting that they made it mandatory to have such a plug in, and the other 70% is just code written before they mandated this.
I would be very surprised if 30% of their code lines had even been touched at all by anyone since AI coding assistants became a thing.
I could see stuff getting small changes and them claiming that the entirety of the new version is “written by AI”.
I wish this shot from The Terminator had the camera showing Sarah Conner’s face instead of Reese’s, because it’d be such an appropriate meme image on multiple levels for when someone makes a misleading claim about some current AI system.
Also, having 1/3 lines with obvious code that can be auto suggested correctly would make sense, but that is hardly code “written by ai” in the way they suggest.
Those are the easy time savings though, the safe easy stuff the developer doesn’t have to worry about anymore. (Giving them time do the gnarly stuff)
It is exactly the opposite, with simple, predictable auto-complete you didn’t have to worry about that anymore, with LLMs you always have to look at it in detail because every little thing could be just plain completely different and wrong.
I can read way faster than I can type though. You still check it, but it’s pretty good as that kind of stuff once you have an example for it to follow.
Reading code is usually orders of magnitude slower than writing code. Sure, typing might be slower than reading but to check if it is what you intended you have to understand it too.
Well, I’m generally very anti-LLM but as a library author in Java it has been very helpful to create lots of similar overloads/methods for different types and filling in the corresponding documentation comments. I’ve already done all the thinking and I just need to check that the overload makes the right call or does the same thing that the other ones do – in that particular case, it’s faster. But if I myself don’t know yet how I’m going to do something, I would never trust an AI to tell me.
Well, okay, I can see how it would be useful in languages like Java that are extremely verbose and have a low expressiveness. Writing Java pretty much was already IDEs with code generation 20 years or so ago because nobody wants to write so much boilerplate by hand.
I’d guess a lot of the people writing the code don’t even have it turned on, it’s just installed because management said it had to be, because management wants to be able to tell investors they’re “innovating work flows”.
Every few months I turn it on for a few days just to see if it is better.
Then I go back to the old AST based autocomplete that actually knows something useful about my code.
I am a small sample to confirm that’s exactly the reason in my brother’s company.
And in my company we’re pressured to make X prompts every week to the company’s own ChatGPT wrapper to show we’re being productive. Even our profit shares have a KPO attached to that now. So many people just type “Hello there” every morning to count as another interaction with the AI.
