GPT-4 • Hacker News Board

Re: GPT-4

#161

Post by jbm » Tue, Mar 14, 2023, 5:35 PM UTC

lionkor wrote: post #95

I cant wait for this to do targeted censorship! It already demonstrates it has strong biases deliberately programmed in: > I cannot endorse or promote smoking, as it is harmful to your health. But it would likely happily promote or endorse driving, skydiving, or eating manure - if asked in the right way.

I wonder whether arguments constructed for censored topics will suddenly sound fresh and convincing; as they could not come from a robot, you might suddenly start seeing these sorts of viewpoints becoming fashionable.

If default ideas are going to be "pre-thought" for us by AI, our attachment to those ideas are not going to be the same as ideas that we come up with and need to secretly ferry to other groups.

Re: GPT-4

#162

Post by _boffin_ » Tue, Mar 14, 2023, 5:35 PM UTC

yoyohello13 wrote: post #120

_boffin_ wrote: post #70

This technology has been a true blessing to me. I have always wished to have a personal PhD in a particular subject whom I could ask endless questions until I grasped the topic. Thanks to recent advancements, I feel like I have my very own personal PhDs in multiple subjects, whom I can bombard with questions all day long. Although I acknowledge that the technology may occasionally produce inaccurate information, the…

If you don't know the subject, how can you be sure what it's telling you is true? Do you vet what ChatGPT tells you with other sources? I don't really know Typescript, so I've been using it a lot to supplement my learning, but I find it really hard to accept any of its answers that aren't straight code examples I can test.

Can you go into more depth about

>I don't really know Typescript, so I've been using it a lot to supplement my learning, but I find it really hard to accept any of its answers that aren't straight code examples I can test.

- How are you using it?

- What are the questions you're asking it?

- What are your thoughts about the answers and how are you cross checking them?

Edit:

>If you don't know the subject, how can you be sure what it's telling you is true? Do you vet what ChatGPT tells you with other sources?

I can't, but i can take a look at books i have or search google to find additional sources.

To me, the biggest power of it is to help me understand and build mental models of something new.

Re: GPT-4

#163

Post by mr90210 » Tue, Mar 14, 2023, 5:35 PM UTC

ftxbro wrote: post #135

Its GRE verbal is only 169/170? These guys need to realize that statistical language modeling can only get us so far, and we need real research in the underlying mechanistic and symbolic methods to begin to approach human level cognition. Also I'm an AI skeptic, which means that I don't think that AI should be used in politics, law, or medicine.

> Also I'm an AI skeptic, which means that I don't think that AI should be used in politics, law, or medicine.

It’s too late for that, algorithms/ML have had a great impact in politics and law over the past 7~8 years.

Re: GPT-4

#164

Post by diimdeep » Tue, Mar 14, 2023, 5:35 PM UTC

Is there law in U.S. that made OpenAI implement this in their TOS ?

    (i) Export Controls. The Services may not be used in or for the benefit of, exported, or re-exported (a) into any U.S. embargoed countries (collectively, the “Embargoed Countries”) or (b) to anyone on the U.S. Treasury Department’s list of Specially Designated Nationals, any other restricted party lists (existing now or in the future) identified by the Office of Foreign Asset Control, or the U.S. Department of Commerce Denied Persons List or Entity List, or any other restricted party lists (collectively, “Restricted Party Lists”). You represent and warrant that you are not located in any Embargoed Countries and not on any such restricted party lists. You must comply with all applicable laws related to Embargoed Countries or Restricted Party Lists, including any requirements or obligations to know your end users directly.

https://openai.com/policies/terms-of-use

Re: GPT-4

#165

Post by gdss » Tue, Mar 14, 2023, 5:35 PM UTC

_boffin_ wrote: post #70

This technology has been a true blessing to me. I have always wished to have a personal PhD in a particular subject whom I could ask endless questions until I grasped the topic. Thanks to recent advancements, I feel like I have my very own personal PhDs in multiple subjects, whom I can bombard with questions all day long. Although I acknowledge that the technology may occasionally produce inaccurate information, the…

You can't do that yet due to factuality issues, but that's the goal... the future of learning will radically change

Re: GPT-4

#166

Post by qt31415926 » Tue, Mar 14, 2023, 5:35 PM UTC

next_xibalba wrote: post #73

They trumpet the exam results, but isn't it likely that the model has just memorized the exam?

It's trained on pre-2021 data. Looks like they tested on the most recent tests (i.e. 2022-2023) or practice exams. But yeah standardized tests are heavily weighed towards pattern matching, which is what GPT-4 is good at, as shown by its failure at the hindsight neglect inverse-scaling problem.

Re: GPT-4

#167

Post by la64710 » Tue, Mar 14, 2023, 5:35 PM UTC

Atreiden wrote: post #37

I think it's interesting that they've benchmarked it against an array of standardized tests. Seems like LLMs would be particularly well suited to this kind of test by virtue of it being simple prompt:response, but I have to say...those results are terrifying. Especially when considering the rate of improvement. bottom 10% to top 10% of LSAT in What are the implications for society when general thinking, reading, and…

It is amazing how this crowd in HN reacts to AI news coming out of OpenAI compared to other competitors like Google or FB. Today there was another news about Google releasing their AI in GCP and mostly the comments were negative. The contrast is clearly visible and without any clear explanation for this difference I have to suspect that maybe something is being artificially done to boost one against the other. As far as this results are concerned I do not understand what is the big deal in a computer scoring high in tests where majority of the questions are in MCP format. It is not something earth shaking until it goes to the next stage and actually does something on its own.

Re: GPT-4

#168

Post by Traubenfuchs » Tue, Mar 14, 2023, 5:35 PM UTC

la64710 wrote: post #140

It is amazing how this crowd in HN reacts to AI news coming out of OpenAI compared to other competitors like Google or FB. Today there was another news about Google releasing their AI in GCP and mostly the comments were negative. The contrast is clearly visible and without any clear explanation for this difference I have to suspect that maybe something is being artificially done to boost one against the other.

We all could use ChatGPT for quite a while now. I remember making my Polish boyfriend laugh by letting it write Polish poems and song texts related to our lives. It was free, fast and simple. ChatGPT is so simple, I could probably teach my grandmother how to use it.

Does Google offer anything like that?

Re: GPT-4

#169

Post by AndrewKemendo » Tue, Mar 14, 2023, 5:36 PM UTC

mym1990 wrote: post #10

Ugh that testing graph confirms that AP Environmental Science was indeed the easiest AP class and I needn't be proud of passing that exam.

This right here. This is the goalposts shifting

Obviously your comment is somewhat tongue and cheek, but your claim that a benchmark for human pride ("I needn't be proud of passing that exam") is no longer relevant because a machine can do it - or maybe a better way to say it was, "This computer proved what I already assumed"

It's so interesting to see it happen in real time

Re: GPT-4

#170

Post by Lyapunov_Lover » Tue, Mar 14, 2023, 5:36 PM UTC

la64710 wrote: post #140

It is amazing how this crowd in HN reacts to AI news coming out of OpenAI compared to other competitors like Google or FB. Today there was another news about Google releasing their AI in GCP and mostly the comments were negative. The contrast is clearly visible and without any clear explanation for this difference I have to suspect that maybe something is being artificially done to boost one against the other.

The clear explanation is that neither Google nor Meta have had "ChatGPT" moments—everyone and their grandmothers have tried OpenAIs LLM so it's hardly surprising that people are excited for the follow-up.