ChatGPT is a blurry JPEG of the web • Hacker News Board

Re: ChatGPT is a blurry JPEG of the web

#61

Post by O__________O » Thu, Feb 09, 2023, 5:15 PM UTC

bbor wrote: post #14

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…

It’s an anti-pattern on HN, that is opinion of its own expertise frequently exceeds reality.

For what it’s worth, specific to the topic at hand, here’s one HN post that provides explanations of how to overcome some of the limitations of large language models:

https://news.ycombinator.com/item?id=34710079

Re: ChatGPT is a blurry JPEG of the web

#62

Post by hjadal » Thu, Feb 09, 2023, 5:15 PM UTC

msla wrote: post #57

donohoe wrote: post #40

Earlier quoted context omitted.

Wholly agree. I worked there many years ago, leading the re-design and re-platform (fun dealing with 90 years of archival content with mixed usage-rights) and paywall implementation (don't hate me, it funds journalism). When you see how the stories get made and how people work there, well, its just amazing.

I notice that the extremists never use paywalls, meaning extremism is allowed to spread unchecked. If respectable newspapers and magazines cared about society, they'd follow suit, and give the extremists some competition.

I think respectable papers would go out of business if they followed suit. They are usually respectable because they dont rely on a outside organisation for funding, but instead rely on creating good journalism people want to pay for.

Re: ChatGPT is a blurry JPEG of the web

#63

Post by rideontime » Thu, Feb 09, 2023, 5:17 PM UTC

ly3xqhl8g9 wrote: post #59

Asked ChatGPT to write a short story, in the style of Ted Chiang, about Ted Chiang thinking that a powerful AI cannot write powerful prose and have something strange happen, ChatGPT said: >>> Ted Chiang sat at his desk, staring at the glowing screen in front of him. He had heard of the new generative text AI that was making waves in the tech world, and he was intrigued. Despite its claims of intelligence, Ted believe…

I'll never understand why people find this amusing.

Re: ChatGPT is a blurry JPEG of the web

#64

Post by buildsjets » Thu, Feb 09, 2023, 5:17 PM UTC

bunbunbun wrote: post #11

mrtksn wrote: post #7

They can always use AI based solutions to unblur the JPEG, like this: https://twitter.com/maxhkw/status/1373063086282739715

As if to say: I heard you like AI outputs, so I put an AI output through another AI so you can have the AI output of an AI’s output

Actually, I HAVE asked ChatGPT to write a prompt for DALL.E 2, and it really did work better that the prompts I wrote myself.

Re: ChatGPT is a blurry JPEG of the web

#65

Post by macintux » Thu, Feb 09, 2023, 5:18 PM UTC

msla wrote: post #57

donohoe wrote: post #40

Earlier quoted context omitted.

Wholly agree. I worked there many years ago, leading the re-design and re-platform (fun dealing with 90 years of archival content with mixed usage-rights) and paywall implementation (don't hate me, it funds journalism). When you see how the stories get made and how people work there, well, its just amazing.

I notice that the extremists never use paywalls, meaning extremism is allowed to spread unchecked. If respectable newspapers and magazines cared about society, they'd follow suit, and give the extremists some competition.

I have yet to see any evidence that “sunlight disinfects” or that mainstream media can out-shout disinformation.

Re: ChatGPT is a blurry JPEG of the web

#66

Post by 13years » Thu, Feb 09, 2023, 5:19 PM UTC

secabeen wrote: post #32

This is a decent summary. I've been thinking about how ChatGPT by it's very nature destroys context and source reputation. When I search for something on the Internet, I get a link to the original content, which I can then evaluate based on my knowledge and the reputation of the original source. Wikipedia is the same, with a big emphasis on citation. ChatGPT and other LLMs destroy that context and knowledge, giving m…

Yes, we need some kind of tool like Explain Plan for databases. Except that for any AI response we can understand more the decision plan and sources.

Re: ChatGPT is a blurry JPEG of the web

#67

Post by iambateman » Thu, Feb 09, 2023, 5:19 PM UTC

To extend the metaphor, blurry jpeg’s have taken us a surprisingly long way as a society.

The trade offs might be alright.

Re: ChatGPT is a blurry JPEG of the web

#68

Post by fnordpiglet » Thu, Feb 09, 2023, 5:19 PM UTC

ElevenLathe wrote: post #18

bbor wrote: post #14

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…

You don't need to correct every wrong thing you read. In fact you will probably feel much better if you don't ever do it at all, or at least take a break for while.

If you did then the LLM might stop hallucinating so much.

Re: ChatGPT is a blurry JPEG of the web

#69

Post by supriyo-biswas » Thu, Feb 09, 2023, 5:20 PM UTC

Any course on information theory typically covers neural network based compression algorithms, so I’m impressed at this observation made by someone who doesn’t have a formal background in CS. Regardless, it’s true.

Re: ChatGPT is a blurry JPEG of the web

#70

Post by trompetenaccoun » Thu, Feb 09, 2023, 5:20 PM UTC

bbor wrote: post #14

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…

I understand why the mainstream thinks that but it's incredibly annoying that even in tech circles there is very little meaningful discussion, it's mainly just people posting amusing screenshots purportedly showing how smart GPT3 is or in other cases how it's politically biased.

Anyone who's played around with it knows that it's fun but it's not a search engine replacement and it doesn't know nor understand things. It regularly gives complete misinformation and when you ask for a source it makes up fake URLs and fake science papers on top of that.

It's nothing like described in the article and I don't understand why people who should know better don't call out the bullshit media reporting more. We've had GPT3 for ages, it's not like most of us only tried it since Chat GPT3 came out, right?