Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Three Years from GPT-3 to Gemini 3

oneusefulthing.org

Post reply on HN
Print view View original

211–220 of 336 posts

Re: Three Years from GPT-3 to Gemini 3

#211

Post by psychoslave »

Madmallard wrote: post #208
psychoslave wrote: post #203

Earlier quoted context omitted.

>You can't be both a genius and extremely dumb (retarded) That’s actually a classic stereotype, someone being a genius in some area, but failing with the most basic social expectations in other area.

In the LLM's case it's both a genius and extremely dumb at coding. The same area.

It’s good at quickly producing a lot of code which is most likely going to give interesting results, and it’s completely unaware of anything including why human might want to produce code.

The marketing bullshit that it’s a "thinking" and "hallucinating" is just bringing the intended confusion on the table.

They are great tools for many purpose. But a GPS is not a copilot, and an LLM is not going to replace coworkers where their humanity matters.

Re: Three Years from GPT-3 to Gemini 3

#212

Post by ManlyBread »

sandspar wrote: post #195
ManlyBread wrote: post #187

Earlier quoted context omitted.

Like clockwork. Each time someone criticizes any aspect of any LLM there's always someone to tell that person they're using the LLM wrong. Perhaps it's time to stop blaming the user?

If someone says that they can't get a camera to work, you tell them how to fix it, right? I can't think of what other response is appropriate.

Why would their response be appropriate when even the creators of the LLM doesn't clearly state the purpose of their software, yet alone instruct users how to use it? The person I replied to said that this software should be used yo "help you build and run experiments, and help you discuss your findings, and in the end helps you write your discoveries" - I dare anyone to find any mention of this workflow being the "correct" way of using any LLM in the LLM's official documentation.

Re: Three Years from GPT-3 to Gemini 3

#213

Post by ragequittah »

ammbauer wrote: post #183
tsss wrote: post #85

Earlier quoted context omitted.

For what it's worth I have been using Gemini 2.5/3 extensively for my masters thesis and it has been a tremendous help. It's done a lot of math for me that I couldn't have done on my own (without days of research), suggested many good approaches to problems that weren't on my mind and helped me explore ideas quickly. When I ask it to generate entire chapters they're never up to my standard but that's mostly an issue…

> It's done a lot of math for me that I couldn't have done on my own (without days of research), Isn't the point of doing the master's thesis that you do the math and research, so that you learn and understand the math and research?

I bet they were talking about how people didn't do long division when the calculator first came out too. Is using matlab and excel ok but AI not? Where do we draw the line with tools?

Re: Three Years from GPT-3 to Gemini 3

#214

Post by matwood »

ManlyBread wrote: post #187
visarga wrote: post #184

Earlier quoted context omitted.

You don't use it that way. You use it to help you build and run experiments, and help you discuss your findings, and in the end helps you write your discoveries. You provide the content, and actual experiments provide the signal.

Like clockwork. Each time someone criticizes any aspect of any LLM there's always someone to tell that person they're using the LLM wrong. Perhaps it's time to stop blaming the user?

You wouldn't use a screwdriver to hammer a nail. Understanding how to use a tool is part of using the tool. It's early days and how to make the best use of these tools is still being discovered. Fortunately a lot of people are experimenting on what works best, so it only takes a little bit of reading to get more consistent results.

Re: Three Years from GPT-3 to Gemini 3

#216

Post by yread »

stavros wrote: post #50
mynameisjody wrote: post #43

Every time I see an article like this, it's always missing --- but is it any good, is it correct? They always show you the part that is impressive - "it walked the tricky tightrope of figuring out what might be an interesting topic and how to execute it with the data it had - one of the hardest things to teach." Then it goes on, "After a couple of vague commands (“build it out more, make it better”) I got a 14 page p…

It's gotten more and more shippable, especially with the latest generation (Codex 5.1, Sonnet 4.5, now Opus 4.5). My metric is "wtfs per line", and it's been decreasing rapidly. My current preference is Codex 5.1 (Sonnet 4.5 as a close second, though it got really dumb today for "some reason"). It's been good to the point where I shipped multiple projects with it without a problem (with eg https://pine.town being one…

I feel it sometimes tries to be overly correct. Like using BigInts when working with offsets in big files in javascript. My files are big but not 53bits of mantissa big. And no file APIs work with bigints. This was from Gemini 3 thinking btw

Re: Three Years from GPT-3 to Gemini 3

#217

Post by yread »

bananaflag wrote: post #41
cm2012 wrote: post #36

Earlier quoted context omitted.

People said the same thing about books and the written word in general

And they were right. Ars memoriae is much less prevalent in the age of mass printed books.

Absolutely. Massive stories were passed on for thousands of year by word of mouth only - all kinds of creation myths. Even Odyssey was oral only for like 300 years, 15 generations!

Re: Three Years from GPT-3 to Gemini 3

#218

Post by gtirloni »

stavros wrote: post #126
gtirloni wrote: post #124

Earlier quoted context omitted.

Maybe the wtfs per line are decreasing because these models aren't saying anything interesting or original.

No, it's because they write correct code. Why would I want interesting code?

Oh, my bad. I still had the comment someone made about the model writing phd-level paper in my head and didn't realize you were talking about code.

Fully agree.

Re: Three Years from GPT-3 to Gemini 3

#219

Post by Madmallard »

psychoslave wrote: post #211
Madmallard wrote: post #208

Earlier quoted context omitted.

In the LLM's case it's both a genius and extremely dumb at coding. The same area.

It’s good at quickly producing a lot of code which is most likely going to give interesting results, and it’s completely unaware of anything including why human might want to produce code. The marketing bullshit that it’s a "thinking" and "hallucinating" is just bringing the intended confusion on the table. They are great tools for many purpose. But a GPS is not a copilot, and an LLM is not going to replace coworkers…

I mean is it really that interesting if it completely falls flat and permanently runs in unfulfilling circles around basically any mild complexity the problem introduces as you get further along solving it, making it really hard to not feel like you need to just do it yourself?

Re: Three Years from GPT-3 to Gemini 3

#220

Post by suddenlybananas »

throwawaylaptop wrote: post #122
taurath wrote: post #96

Earlier quoted context omitted.

But holy shit is it also a funding issue when teachers make nothing.

I date a lot of teachers. My last one was in the San Ramon (CA) Valley School district, she makes about $90k a year at 34 years old. Talking to her basically makes me want to homeschool my kids to make sure someone like her isn't their teacher. Paying teachers more won't do ANYTHING until we become a lot more selective about who gets to become and stay a teacher. It can't be like most government jobs where getting it…

If teachers made as much as half the people on this site, perhaps things would be better. 90k in San Ramon is more or less the median wage [1]. It's not _that_ much money.

[1] https://en.wikipedia.org/wiki/San_Ramon,_California#2020_cen...

Post reply on HN