Imnimo

18 ms·

271.
by Imnimo 3y ago
I'm sympathetic to the idea that training is fair use. I'm much less sympathetic to the "it's a bug" defense on regurgitation. Ultimately, if your system is reproducing copyrighted material, it seems like you'r
272.
by Imnimo 3y ago
>The 24-page list of artists’ names used by Midjourney as the training foundation for its AI image generator I don't see how they reached the conclusion that this list is "the training foundation" of Midjourney. I feel lik
273.
by Imnimo 3y ago
I'd be curious to see if, in the 1-2% of cases where the linear probe fails to predict board occupancy, the LLM also predicts (or at least assigns non-trivial probability to) a corresponding illegal move. For example, if the linear pro
274.
by Imnimo 3y ago
I'm not clear on what the "laundering" is supposed to be here. I feel like that screenshot doesn't give me any context for what the user is saying. Given that Codex is a code model, it seems unlikely it's directly r
275.
by Imnimo 3y ago
This looks a lot more convincing to me than the Copilot lawsuit or the Sarah Silverman one. This suit shows ChatGPT reciting large amounts of NYT articles - not just little snippets of code or the ability to answer questions about Silverman
276.
by Imnimo 3y ago
I disagree that the type of coding the LLM is doing here is "expert" level in any meaningful sense. Look for example at the code for the bin-packing heuristics: https://github.com/google-deepmind/funsearch&#x2
277.
by Imnimo 3y ago
Critically, though, the LLM is not acting as a domain expert here. It's acting as a code mutator. The expertise it brings to the table is not mathematical - Codey doesn't have any special insight into bin packing heuristics. It&#x
278.
by Imnimo 3y ago
I don't think five runs is sufficient to reach the conclusion you're reaching here, especially when we can see that one of the Codey runs is worse than all of the StarCoder runs. If Codey is so much more valuable, why is this happ
279.
by Imnimo 3y ago
Do you think the role the LLM plays in this system is analogous to what Tao is talking about?
280.
by Imnimo 3y ago
>It doesn't take a genius however to see who the more valuable contributor to the system is. Indeed, we can look at the ablations and see that swapping out a weaker code LLM (540B Codey to 15B Starcoder) makes little difference: &qu
281.
by Imnimo 3y ago
We can see from the ablations that simply asking the LLM to generate a large number of potential solutions without the evolutionary search performs terribly. It's certainly true that the LLM mutator produces more reasonable edits than
282.
by Imnimo 3y ago
The heavy lifting here is done by the evolutionary algorithm. The LLM is just being asked "propose some reasonable edits to these 20 lines of python" to replace a random mutation operator. It feels generous to credit the neural ne
283.
by Imnimo 3y ago
Just from the title I knew exactly which awardees we were talking about, and who was going to be doing the talking.
284.
by Imnimo 3y ago
The tweet seems to contradict the screenshot.
285.
by Imnimo 3y ago
Well at least, 23andMe promises that it also can't participate in a class-action lawsuit against me. So that's pretty fair.
286.
by Imnimo 3y ago
This feels like one of those things where a UI designer tries to make the most comically inefficient way to convey mundane information.
287.
by Imnimo 3y ago
I would disagree that that constitutes a "fundamental advantage"
288.
by Imnimo 3y ago
It really feels like the reason this is being released now and not months ago is that that's how long it took them to figure out the convoluted combination of different evaluation procedures to beat GPT-4 on the various benchmarks.
289.
by Imnimo 3y ago
I'm surprised it doesn't take into account how much money Elon thinks you'll be able to make renting it out as a robotaxi.
290.
by Imnimo 3y ago
It feels like 10M is a fraction of what it would cost to train such a model. Even if we assume an eventual winner would be willing to release the model and forgo potential profits, does this prize really motivate development if it doesn&#x2
291.
by Imnimo 3y ago
Quoting Emmett Shear: >The board did not remove Sam over any specific disagreement on safety, their reasoning was completely different from that. So why is this article titled "OpenAI's board had safety concerns"? It doe
292.
by Imnimo 3y ago
So the complaint is that, in order to see if Twitter would place ads next to hate content, Media Matters first followed accounts that post hate and then scrolled down?
293.
by Imnimo 3y ago
What I still don't get about all this is if this is just a case of Ilya deciding to play board room Game of Thrones because he has a different vision for OpenAI than Sam and Greg, why throw Sam under the bus with this line in the annou
294.
by Imnimo 3y ago
People don't know who Ilya is?
295.
by Imnimo 3y ago
>If you start a query with “Mary Lee Pfeiffer”, you’re not going to get very far because neural networks aren’t equidistant grids of points (besides the fact that she may not appear very often under that version of her name.) They’re net
296.
by Imnimo 3y ago
If this isn't just luck, and it's based public and obvious information, why wouldn't the effect already be priced in by large firms trading on it?
297.
by Imnimo 3y ago
>Grok is designed to answer questions with a bit of wit and has a rebellious streak, so please don’t use it if you hate humor! This is the most embarrassing line I have ever read in a product announcement.
298.
by Imnimo 3y ago
So is the way this works that Google pays a bunch of money to ICANN, and then they get the .ing TLD, and can make money by selling individual .ing domains? Is there like a bidding process that ICANN uses to decided who gets the TLD in the f
299.
by Imnimo 3y ago
My guess is that "g" here is primarily the ability of the model to recognize and respond to multiple choice questions (the "performance" here is almost entirely measured on MMLU). Weaker models often do things like recit
300.
by Imnimo 3y ago
I'm curious if this is actually a stable state given their debt situation. Are they actually treading water or are they slowly sinking?

More ›