ElFitz

34 ms·

571.
by ElFitz 4y ago
I might be wrong but I think it’s more of a "It’s obvious you don’t need to have fingers to build a hand that has fingers" than building a fingerless functional equivalent to a hand.
572.
by ElFitz 4y ago
Well, you could use both. One model for content generation, one for moderation and censorship. But, in this case, as already discussed in other threads, why not simply transform everything people write or say? They still get to see what the
573.
by ElFitz 4y ago
If they were to train their own, wouldn’t they only train it on "approved" content? In which case, it wouldn’t even need to be censored, would it?
574.
by ElFitz 4y ago
I also see them as a way of filtering content. Sure, it could lead one to miss out on some relevant things. But there is just so much out there. Most of it mostly noise. Most of the rest, irrelevant to us individually. Being able to navigat
575.
by ElFitz 4y ago
For some reason, the phrasing, the structure, the rhythm, they all very strongly remind me of Alan Watts’. Quite amusing and surprising.
576.
by ElFitz 4y ago
That’s one of the issues with summaries. Not only are some written by writers who couldn’t write as well as the original author, but you also lose information, no matter what you do. Sometimes it’s just the form. Sometimes it’s fluff. But s
577.
by ElFitz 4y ago
For that I'd suggest using Langchain with Wolfram Alpha. It's already been done and discussed: - https://news.ycombinator.com/item?id=34422122 - https://news.ycombinator.com/item?id=34422627
578.
by ElFitz 4y ago
Or using homomorphic encryption. I remember some managing to run inference on encrypted images. See - https://www.zama.ai/post/encrypted-image-filtering-using-hom... - https://news.ycombinator.com/item?
579.
by ElFitz 4y ago
Except for things we decide to be important enough that we should create and apply models. Now, the decisions that went into these models might have been rationalised after the fact. Or biased. But these handmade models can been reviewed by
580.
by ElFitz 4y ago
Depending on the book you could use that to extract all excerpts where the character appears. Then each excerpt could be fed to the LLM asking it how this part relates to the question you’d want an answer to. Then ask for each what it shows
581.
by ElFitz 4y ago
Could using GPT3 (davinci-003) to generate embeddings, then searching your vector database for relevant excerpts, then providing the results as context for the prompt lead to something close enough?
582.
by ElFitz 4y ago
We could have handcrafted stories and quests, with LLM-driven dialogues for NPCs canned responses (ie the infamous arrow and the proverbial knee). And teams with limited resources could also still handcraft the stories and quests but use LL
583.
by ElFitz 4y ago
Oh, I’m definitely with you on that. It’s just that all around the world American Cheese is now often, although wrongly, mistaken for cheddar. And most people have no idea what actual cheddar looks or tastes like. Which is precisely what th
584.
by ElFitz 4y ago
You are quite frankly proving my point: that it is, first and foremost, a branding matter.
585.
by ElFitz 4y ago
I find it amusing because I did spend quite some time trying to figure out how to get the QR code from the app. But that’s because I usually don’t use Amazon US, and you need to change to the US store, in app, to get that qr code. I feel li
586.
by ElFitz 4y ago
Cameras and other sensors. You do need to scan a QR code from the app before coming in. Some other comments mention having trouble with it, being billed for things they shouldn’t have. But I personally never did. Even though I enjoyed fooli
587.
by ElFitz 4y ago
So… Amazon?
588.
by ElFitz 4y ago
Amazon Go was even more incredible than that. Walking out without even having to deal with those pesky self-service cash register, their sometimes finicky scales, and some supermarkets’ more than annoying rules, feels like a weird magic tri
589.
by ElFitz 4y ago
It’s not a matter of using the language properly; it’s a branding matter. It also isn’t just about the origin. "Gruyere", "Cantal" or "Champagne" aren’t only about where the thing is made, but also how it is ma
590.
by ElFitz 4y ago
Generative AI could be the perfect tool for a 1984 world. Remove whatever is inconvenient from the record, replace it with 10x more generated ahistorical content. At nearly 0 marginal cost.
591.
by ElFitz 4y ago
That is good to know. Thanks!
592.
by ElFitz 4y ago
> it seems like nothing has really replicated Planetside 2's black magic netcode Or tried to. From what I’ve read online they massively rely on client-side computation, even for hit detection. It seems more like trading security for
593.
by ElFitz 4y ago
Is there an API for these things? It could be a good way of easily and automatically letting all those that rely on their services let their own customers know there is an issue.
594.
by ElFitz 4y ago
There's also Azure's. That's if you manage to signup and are allowed to even apply for access to their OpenAI services.
595.
by ElFitz 4y ago
> it can't be original, it will forever recycle/regurgitate past SF I can’t see why. It could very well, on its own, use other genre’s narrative structures, stories or styles that are uncommon or inexistant in SF. It could also
596.
by ElFitz 4y ago
And once again, we prove it: there is an XKCD for everything (although I have to admit this one was easy). https://xkcd.com/927/
597.
by ElFitz 4y ago
The analogy doesn’t hold. ´ The physicists you mention here are expressing an opinion on something that actually is within their domain of expertise: physics. I fail to see how psychology, sociology, economics or politics fall within an AI
598.
by ElFitz 4y ago
Well, that should be easy!
599.
by ElFitz 4y ago
On that topic, I am more and more starting to believe that most of the "aesthetic" aspect of code style (ie types of quotes, wether or not braces should be on a new line, spacing, spacing type, line length, etc…) should be simply
600.
by ElFitz 4y ago
It definitely does. But if I'm never going to be able to get feedback of any sort on them, or even know if anyone read it, why should I bother with hosting and maintaining it online? This use case can be solved for using a pen and pape

More ›