Look at it, Doodle AG Scheduling Software Pro Annual Subscription per User, and weep
Ye, so essentially a wireless Avada Kedavra, cool cool cool, completely chill and sane thing to believe
195 IQ and suddenly get someone who just sits in their room for a decade and then speaks gibberish into a youtube livestream and everyone dies, or whatever.
I can't even decipher what this is about. Like if you're 195IQ you can invent Avada Kedavra in a decade?
Artificial wombs may remove this bottleneck.
Okay but this is an amazing out-of-context sentence. I will croudfund a $1000 award for anyone who is able to put that sentence into a paper and get published in Nature without anyone noticing.
I don't think Harry was much of a genius, unless you mean Harriezer from MoR in which case lol, lmao
Working in the field of genetics is a bizarre experience
How the fuck would you know that, mate? You don't even have a degree in your field, which, let me remind you, is (allegedly) computer science. Has Yud ever been near an actual genetics professor?
I feel coding people like they’re software
Jesus christ can you imagine segfaulting someone's kidney
It's reacting to the presentation, not you specifically. I think many of the other comments hit on how he goes waaay too far in his criticism, but I wouldn't have written what I wrote if it wasn't a wider sentiment I encountered a few times already.
The attitude to theoretical computer science re quantum is really weird. Some people act as if "I can't run it now therefore it's garbage" which is just such a nonsense approach to any kind of theoretical work.
Turing wrote his seminal paper in 1936, over 10 years before we invented transistors. Most of CS theory was developed way before computers were proliferated. A lot of research into ML was done way before we had enough data and computational power to actually run e.g. neural networks.
Theoretical CS doesn't need to be recent, it doesn't need to run, and it's not shackled to the current engineering state of the art, and all of that is good and by design. Let the theoreticians write their fucking theorems. No one writing a theoretical paper makes any kinds of promises that the described algorithm will EVER be run on anything. Quantum complexity theory, for example was developed in the nineties, there was NO quantum computer then, no one was even envisioning a quantum computation happening in physical reality. Shor's algorithm was devised BEFORE THAT, before we even had the necessary tools to describe its complexity.
I find the line of argumentation "this is worthless because we don't know a quantum computer is engineeringly feasible"
The reason is that any government mandated ID is clearly the Mark of the Beast and will be used to bring upon a thousand years of darkness.
You think that's fringe nonsense and you'd be right on the nonsense part, but that's literally what Ronny Reagan said while he was president
1970s probably?
Turns out software engineering cannot be easily solved with a ~~small shell script~~ large language model.
The author of the article appears to be a genuine ML engineer, although some of his takes aged like fine milk. He seems to be shilling Google a bit too much for my taste. However, the sneer content is good nonetheless.
First off, the "Devin solves a task on Upwork" demo is 1. cherry picked, 2. not even correctly solved.
Second, and this is the absolutely fantastic golden nugget here, to show off its "bug solving capability" it creates its own nonsensical bugs and then reverses them. It's the ideal corporate worker, able to appear busy by creating useless work for itself out of thin air.
It also takes over 6 hours to perform this task, which would be reasonable for an experienced software engineer, but an experienced software engineer's workflow doesn't include burning a small nuclear explosion worth of energy while coding and then not actually solving the task. We don't drink that much coffee.
The next demo is a bait-and-switch again. In this case I think the author of the article fails to sneer quite as much as it's worthy -- the task the AI solves is writing test cases for finding the Least Common Multiple modulo a number. Come on, that task is fucking trivial, all those tests are oneliners! It's famously much easier to verify modulo arithmetic than it is to actually compute it. And it takes the AI an hour to do it!
It is a bit refreshing though that it didn't turn out DEVIN is just Dinesh, Eesha, Vikram, Ishani, and Niranjan working for $2/h from a slum in India.
It's infinite monkeys but every time they output coherent English you give them bananas to incentivise them towards that