111–120 of 317 posts
Earlier quoted context omitted.
If this article is the best the New Yorker offers now, I'm glad I don't subscribe. It used to have high-quality articles, certainly.
I thought the author was uncharacteristically perceptive for a reporter. Yann LeCun or Geoff Hinton couldn't have come up with a better analogy.
I was remembering a quote too vaguely to find the original with Google. I explained the idea of the quote to ChatGPT and it pointed me directly to the quote in its original language and its author. I could then easily look it up on Google.
Earlier quoted context omitted.
It's just the potential. Something feels in the right direction. Perhaps it will only remain a tool and it will never write the next Infinite Jest , but the next David Foster Wallace will certainly use its potential to become a better writer, just in the same manner auto-spellcheckers make me a better writer. Incidentally, David Foster Wallace would have loved ChatGPT: finally, he no longer needed to write, he could…
> but the next David Foster Wallace will certainly use its potential to become a better writer, just in the same manner auto-spellcheckers make me a better writer. Really contestable, actually, not certain. This is also addressed in the text: > Can large-language models help humans with the creation of original writing? To answer that, we need to be specific about what we mean by that question. There is a genre of ar…
So, yes, ChatGPT will probably not be essential in the creation of art, even today you could write a great novel with merely $10 for pencil and paper, but it will be better with ChatGPT or similar tools: faster to find the word, easier to iterate over the possibilities of a phrase, helpful to evaluate parallel scenarios, and other use cases to be found by great writers.
Sure, there might be an issue for the artists themselves: who do you reward, is it "valid" art, and so forth. But as a reader, ChatGPT hints that the greatest works of literature are yet to come: hundreds of times while reading a page I would have wanted to read a hundred more pages about a particular aspect but the author went frustratingly in another direction. In this case, the author might be elevated, from the laborer putting words on paper, to a generator of directions to be followed by the generator of text, in the same manner Michelangelo, Caravaggio, or Gaudi directed the painters, sculptors, and workers to execute accordingly.
Tautologically, yes, ChatGPT works because it is, as defined by the author, a lossy algorithm. If it were a lossless algorithm it wouldn't work the way it does now.
> The fact that ChatGPT rephrases material from the Web instead of quoting it word for word makes it seem like a student expressing ideas in her own words, rather than simply regurgitating what she’s read; it creates the illusion that ChatGPT understands the material. In human students, rote memorization isn’t an indicator of genuine learning, so ChatGPT’s inability to produce exact quotes from Web pages is precisely what makes us think that it has learned something. When we’re dealing with sequences of words, lossy compression looks smarter than lossless compression.
This is where the analogy of a lossy and lossless compression algorithm breaks down. Yes, a loosely similar approach of principle component analysis and dimensional reduction as used in lossy compression algorithms is being applied and we can see that most directly in a technical sense with GPT `embedding vector(1536)`, but there is a big difference: ChatGPT is also a translator and not just a synthesizer.
This has nothing to do with "looking smarter". It has to do with being reliably proficient at both translating and synthesizing.
When given an analytic prompt like "turn this provided box score into an entertaining outline", ChatGPT proves itself to be a reliable translator, because it can reference all of the facts in the prompt itself.
When given a synthetic prompt like "give me some quotes from the broadcast", ChatGPT proves itself to be a reliable synthesizer, because it can provide fictional quotes that sound correct when the facts are not present in the prompt itself.
The synthetic prompts function in a similar manner to lossy compression algorithms. The analytic prompts do not. This lossy compression algorithm theory, also known as the bullshit generator theory, is an incomplete description of large language models.
https://williamcotton.com/articles/chatgpt-and-the-analytic-...
This is very well written, and probably one of my favorite takes on the whole ChatGPT thing. This sentence in particular: > Indeed, a useful criterion for gauging a large-language model’s quality might be the willingness of a company to use the text that it generates as training material for a new model. It seems obvious that future GPTs should not be trained on the current GPT's output, just as future DALL-Es should…
> the recursive feedback loop would just yield nonsense an assumption disguised as fact. we simply do not know yet
Earlier quoted context omitted.
Analogy of?
"Blurry JPEG" for how ChatGPT "compresses" character-based knowledge into vectors. That "compression" process gives ChatGPT an ability to generalize because it learns statistics (unlike JPEG) but like JPEG it is a lossy process.
Earlier quoted context omitted.
> This is like when my father excitedly told his friends about the coming computer revolution in the 90s and they responded “well it can’t do my dishes or clean the house, they’re just a fad!” Makes me want screaaaaam Did the computer revolution make your father’s life better? Serious question.
Hmm. I’d say definitely yes. I mean we are on the internet right now, presumably across many km of distance. Do you disagree?
Earlier quoted context omitted.
I’m just searching the comments for novel use cases where its effective. Most articles I’ve read seem like either moral panics or snake oil. I like how it can generate songs and poems based on a prompt. Its not particularly useful, but it is entertaining. It really does seem curated at times, leading me to think this will eventually become a fad or replaced by a more advanced algorithm.
The most potent use case of the current implementation of ChatGPT is plagiarism. Businesses who need blog posts for content marketing would hire copywriters who would paraphrase my content and publish it as their own. Now they can do it faster and for free. Blogspam got cheaper and more effective. This applies to other kinds of spam too: email spam, forum spam, comment spam... Spam spam spam. I'm afraid that we'll so…
Not sure how to phrase an email in a professional tone, tell GPT to write it and edit from there. This is a tool that is going to be a smash it with non-native speakers or really anyone who just isn't that good with words.
> Write an email to a coworker telling them in the politest and most professional manner that they're a fucking idiot who doesn't know what they're talking about and that I'm the subject matter expert on this topic. Don't swear and have a firm but neutral tone.
I understand that you have a great interest in the topic we were discussing earlier, and I appreciate your enthusiasm. However, I feel it is important to clarify a few points.
As the subject matter expert on this topic, I believe it is important to note that there are certain aspects that may not be fully understood without a deeper understanding of the material. With that being said, I would like to offer my assistance in making sure that all aspects of the topic are properly understood.
If you have any further questions or would like to discuss the topic in more detail, I am more than happy to assist. Please do not hesitate to reach out at any time.
Ugh I’m beginning to think I’m going to spend the next 6-12 months commenting “no, large language models aren’t supposed to somehow know everything in the world. No, that’s not what they’re designed for. Yes, hooking one up to our long-standing record-of-everything-in-the-world (google’s knowledge graph) is going to be powerful.” It’s getting to point where I need to consider stop going on HN. This is like when my fa…
> Yes, hooking one up to our long-standing record-of-everything-in-the-world (google’s knowledge graph) is going to be powerful. This hasn't happened yet and, while I may just lack imagination, despite having a fairly solid understanding of how the latest round of AI works I can't see how it can be done successfully. Until it is in fact done and my lack of imagination is demonstrated, your "going to be powerful" is a…
Earlier quoted context omitted.
> well it can’t do my dishes or clean the house the correct response would have been "yet"
Well, 30 years later and they still can't. At least I know of no one who has a house cleaning robot. (I wouldn't classify Roomba as 'clean my house').