No no I am talking of actual non bullshit work on the underlying math. Think layernorm, skip connections, that sort of thing, changes how the neural network is computed so that it trains more effectively. edit: in that case would be changing it so that after training, at inference for the typical query, most (intermediary) values computed will be zero.
Now we need to make a logic puzzle involving two people and one cup. Perhaps they are trying to share a drink equitably. Each time they drink one third of remaining cup’s volume.
I appreciate the sentiment but I also hate the whole "AI is a power loom for coding".
The power loom for coding is called "git clone".
What "AI" (LLM) tools provide is just English as a programming language with plagiarized sum total of all open source as the standard library. English is a shit programming language. LLMs are shit at compiling it. Open source is awesome. Plagiarized open source is "meh" - you can not apply upstream patches.
Old McDonald had a startup, iyo io o[4-mini].
It's funny how just today in a completely unrelated context a generative ai enthusiast used an example of OpenAI getting sued by NYT as a reason why they wouldn't commit some other malfeasance because they'd get caught if they did.
Having worked in computer graphics myself, it is spot on that this shit is uncontrollable.
I think the reason is fundamental - if you could control it more you would put it too far from any of the training samples.
That being said video enhancements along the lines of applying this as a filter to 3d rendered CGI or another video, that could (to some extent) work. I think the perception of realism will fade as it gets more familiar - it is pretty bad at lighting, but in a new way.
Oh and also for the benefit of our AI fanboys who can't understand why we would expect something as mundane from this upcoming super-intelligence, as doing math, here's why:
Incels then: Zuckerberg creates a hot-or-not clone with stolen student data, gets away with it, becomes a billionaire.
Incels now: chatgpt, what's her BMI.
I am also presuming this is about purely non-fiction technical books
He has Dune on his list of worlds to live in, though...
edit: I know. he fed his post to AI and asked it to list the fictional universes he'd want to live in, and that's how he got Dune. Precisely the information he needed.
is somewhere between 0 and 100%.
That really pins it down, doesn't it?
Yeah. I’d love to see the prompt, gab’s nazi ai prompt was utterly pathetic and this one got to be pretty bad as well.
Yeah I think the best examples are everyday problems that people solve all the time but don't explicitly write out solutions step by step for, or not in the puzzle-answer form.
It's not even a novel problem at all, I'm sure there's even a plenty of descriptions of solutions to it as part of stories and such. Just not as "logical puzzles" due to triviality.
What really annoys me is when they claim high performance on benchmarks consisting of fairly difficult problems. This is basically fraud, since they know full well it is still entirely "knowledge" reliant, and even take steps to augment it with generated problems and solutions.
I guess the big sell is that it could use bits and pieces of logic gleaned from other solutions to solve a "new" problem. Except it can not.
diz
joined 2 years ago
I think provenance has value outside copyright... here's a hypothetical scenario:
libsomeshit is licensed under MIT-0 . It does not even need attribution. Version 3.0 has introduced a security exploit. It has been fixed in version 6.23 and widely reported.
A plagiaristic LLM with training date cutoff before 6.23 can just shit out the exploit in question, even though it already has been fixed.
A less plagiaristic LLM could RAG in the current version of libsomeshit and perhaps avoid introducing the exploit and update the BOM with a reference to "libsomeshit 6.23" so that when version 6.934 fixes some other big bad exploit an automated tool could raise an alarm.
Better yet, it could actually add a proper dependency instead of cut and pasting things.
And it would not need to store libsomeshit inside its weights (which is extremely expensive) at the same fidelity. It just needs to be able to shit out a vector database's key.
I think the market right now is far too distorted by idiots with money trying to build the robot god. Code plagiarism is an integral part of it, because it makes the LLM appear closer to singularity (it can write code for itself! it is gonna recursively self-improve!).