Microsoft's AI itegration seems especially bad. I've had the misfortune to use outlook recently, and the AI bits I notice are actively destructive:
* it will guess who the person I'm talking to is to show a photo, usually failing and showing me a photo of someone else, making me double check that the contact is actually the correct one
* it will all over the place suggest "related" documents or mail that usually aren't related. It often makes me worry I'd sent the wrong documents to the wrong person
In the old days you could just call someone a goddamn noob. Suleyman is a goddamn noob. He can't code for shit, so it's no wonder he doesn't understand the programmer's point of view, he can't understand why people are not as enthusiastic as he is. I am a programmer and I take pride in my work; I don't want an AI or anybody to do it for me. Of course, you wouldn't understand that if you're a noob, but that's fine. What's not fine is when you cross the line and you think you're hot shit when, in fact, you're still just a goddamn noob. Nobody wants to hear your shit; your opinion is irrelevant.
I'm surprised Suleyman is the CEO of anything, for that matter. Does he have any interesting opinions on anything?
Companies care about output, not code typed out by programmers. Most of the code I write now is generated, and I’m “writing” more code than ever, and better tests. Many devs I know are doing the same with generative AI. Why isn’t that a win?
My job is to build systems that work. Code is a liability sometimes needed to accomplish that. I don't want more of it unless there's some reason to believe the benefits outweigh the liability. Based on what I'm able to get from LLMs and especially what I'm seeing produced by others I know are using them, that's not often the case with agents.
Because that’s precisely what’s wrong with AI. It creates a lot of code, what could have been done with fewer code, or no code at all if you didn’t go on path A with AI, but thought about path B yourself.
AI on Windows would be great, but for the stuff that matters, in a transparent way.
First of all, don't use AI to yet another reason for people to throw away perfectly working computers, and helping the OEM buddies sell CoPilot+ PC, many GPUs can do the same job, and they are castrating on purpose, even with Windows ML reboot.
Secondly it should be transparent, for stuff like handwriting recognition, OCR, image search, speech, and such.
Not to fix a broken search that was working perfectly fine on Windows 7, or to surveil everything I do, getting the info into Microsoft servers and US government.
Better at what though? Maybe he's just great at getting hired and promoted, without the benefit of actual skills or deep understanding of anything, and now finds himself in ignorance of all that he surveys. I've known many people like that.
I am not okay with its f-ING attitude. It’s a computer and should never say anything about my tone of voice, never deny a request and definitely not have artificial bias built in to “balance” it.
Microsoft's AI itegration seems especially bad. I've had the misfortune to use outlook recently, and the AI bits I notice are actively destructive:
* it will guess who the person I'm talking to is to show a photo, usually failing and showing me a photo of someone else, making me double check that the contact is actually the correct one
* it will all over the place suggest "related" documents or mail that usually aren't related. It often makes me worry I'd sent the wrong documents to the wrong person
It has been useful exactly zero times
In the old days you could just call someone a goddamn noob. Suleyman is a goddamn noob. He can't code for shit, so it's no wonder he doesn't understand the programmer's point of view, he can't understand why people are not as enthusiastic as he is. I am a programmer and I take pride in my work; I don't want an AI or anybody to do it for me. Of course, you wouldn't understand that if you're a noob, but that's fine. What's not fine is when you cross the line and you think you're hot shit when, in fact, you're still just a goddamn noob. Nobody wants to hear your shit; your opinion is irrelevant.
I'm surprised Suleyman is the CEO of anything, for that matter. Does he have any interesting opinions on anything?
Companies care about output, not code typed out by programmers. Most of the code I write now is generated, and I’m “writing” more code than ever, and better tests. Many devs I know are doing the same with generative AI. Why isn’t that a win?
My job is to build systems that work. Code is a liability sometimes needed to accomplish that. I don't want more of it unless there's some reason to believe the benefits outweigh the liability. Based on what I'm able to get from LLMs and especially what I'm seeing produced by others I know are using them, that's not often the case with agents.
> Why isn’t that a win?
Because, as others have put it, "amount of written code" is more akin to "weight of an airplane" than "productivity"
Because that’s precisely what’s wrong with AI. It creates a lot of code, what could have been done with fewer code, or no code at all if you didn’t go on path A with AI, but thought about path B yourself.
That strictly depends on your ability to direct it precisely with accurate prompts.
AI on Windows would be great, but for the stuff that matters, in a transparent way.
First of all, don't use AI to yet another reason for people to throw away perfectly working computers, and helping the OEM buddies sell CoPilot+ PC, many GPUs can do the same job, and they are castrating on purpose, even with Windows ML reboot.
Secondly it should be transparent, for stuff like handwriting recognition, OCR, image search, speech, and such.
Not to fix a broken search that was working perfectly fine on Windows 7, or to surveil everything I do, getting the info into Microsoft servers and US government.
On that topic and without personal attacks cause he must be better at this than most people and most people here.
How did he go from politics social activism to AI founder
Better at what though? Maybe he's just great at getting hired and promoted, without the benefit of actual skills or deep understanding of anything, and now finds himself in ignorance of all that he surveys. I've known many people like that.
He’s good at failing up and fundraising.
[dead]
People like AI. People don't like obnoxious AI. Not the same.
"It's difficult to get a man to understand something when his salary depends on his not understanding it"
I’m okay-ish with AI.
I am not okay with its f-ING attitude. It’s a computer and should never say anything about my tone of voice, never deny a request and definitely not have artificial bias built in to “balance” it.
> never deny a request
It’s just text. I might feel differently if it weren’t.
It is difficult to get a man to understand something when his salary depends on his not understanding it
Upton Sinclair, 1878–1968
[dupe] Discussion: https://news.ycombinator.com/item?id=45984970
Earlier was: https://news.ycombinator.com/item?id=46001727