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

CheatGPT

blog.humphd.org

Post reply on HN
Print view View original

501–510 of 544 posts

Re: CheatGPT

#501

Post by sanderjd »

grugagag wrote: post #446
sanderjd wrote: post #289

Earlier quoted context omitted.

In my last round of interviews, by far my favorite session was "here is a real project with a real bug, debug it, then if you have time, describe how to fix it, then if you have time, write the fix". (At Stripe FWIW.) It was super representative of what my work is actually like, and what I'm good at, and ChatGPT would not have figured it out. yet

That sounds very fair. I’d prefer that over leetcode type of questions anytime

Yeah! I didn't end up taking a job at Stripe, but I came away very impressed with their process. Only one interview was leetcode-ish and it was a phone screen and purposefully not a high bar. Then the real interviews were: that debugging one which I loved, an "integration" one which involved live coding while plugging a few different unfamiliar libraries and APIs together (also something I really do in my job), a standard system design interview (still don't really love these...), and an in depth conversation with a hiring manager. Still a tough day, but pretty decent process I think!

Re: CheatGPT

#503

Post by sebastianconcpt »

The problem with synthetic intelligence usage is that, if used like a crutch, it will necessarily modify how you walk. And if you could develop walking just fine and interject this, are you going to be a great walker?

What would be the next generation's intelligence strong at?

Biology needs a soft quote of regular stress to develop compensating the needed optimizations. Happens with muscle growth, fat burning and developing psychological traits and behavioral transformations.

What will be our new opportunities for growth if all the hard work is going to be done by synthetics? (with a hardware to run that is as expensive as a high category car)

Re: CheatGPT

#504

Post by Jensson »

sebastianconcpt wrote: post #503

The problem with synthetic intelligence usage is that, if used like a crutch, it will necessarily modify how you walk. And if you could develop walking just fine and interject this, are you going to be a great walker? What would be the next generation's intelligence strong at? Biology needs a soft quote of regular stress to develop compensating the needed optimizations. Happens with muscle growth, fat burning and dev…

Intelligence will go the same way bodyweight did when we stopped needing to use our bodies. Some people exercise and eat properly still, but most doesn't.

Re: CheatGPT

#505

Post by jraph »

Al-Khwarizmi wrote: post #490
jraph wrote: post #316

Earlier quoted context omitted.

> the continuous assessment fad that has always been rather disastrous for equity Can you expand on this? I lived the two models as a student and still can't make up my mind on which is best. I feel like it just takes catching some bad cold the week of the exams or a bad exam subject / topic to fail them which seems pretty unfair, but I also know my shit better and have more perspective the week of the exam and prefe…

This is probably an unpopular opinion, but in my view, traditional final exams tend to be a great equalizer. Continuous assessment is affected much more by personal circumstances. For example, continuous assessment tends to force students to devote a constant amount of time throughout the course, which is fine for those who can afford to be full-time students, but disastrous for students who need to have a job to pay…

Okay, totally makes sense. Thanks!

I wish people didn't have to work during studies. That's one of the biggest reasons of failing.

Re: CheatGPT

#506

Post by Sophira »

consumer451 wrote: post #419
Sophira wrote: post #416

Earlier quoted context omitted.

As a control, I tried asking the same question but asking about volume instead. Unfortunately, ChatGPT appears to get it wrong: User: > Which is heavier: one cubic foot of nails, or one cubic foot of feathers? ChatGPT: > One cubic foot of nails and one cubic foot of feathers would weigh the same amount. This is because weight is determined by the amount of matter in an object, which is measured by its mass, and not i…

Great follow up! Well this is very interesting. I pasted your prompt into a new chat just now and I got the correct answer: > One cubic foot of nails and one cubic foot of feathers both have the same volume, but nails are denser than feathers. Therefore, one cubic foot of nails will be heavier than one cubic foot of feathers. Are you using the free version at chat.openai.com? Does it say Feb 13 version at the bottom?

I am, yes.

However, I did initially misspell the prompt to say "heaver", and with the misspelling it had answered correctly. Once I corrected it to "heavier", it then came up with the wrong answer.

Looking at it with fresh eyes, it's possible that my methodology was flawed. While I only ever intended to showcase the first response, I did also try regenerating the responses a few times. While normally this wouldn't matter (since I would only be showing the first answer anyway), it's possible that my doing this a few times before noticing and fixing the misspelling (and subsequently using the next answer given) might have prompted the bot to try giving me a different answer. (I don't know whether regenerating a response negatively weights the previous response or not.)

I apologise for not mentioning this! In my defense, I was tired.

Re: CheatGPT

#507

Post by ryanjshaw »

esperent wrote: post #496
ryanjshaw wrote: post #479

Earlier quoted context omitted.

I fix plenty of weird bugs. But I don't build a reverse index in my mind of "weird bugs".

Can you think of any bug from that last five months that you could talk about for five minutes in a way that makes you look smart for solving it? That's all you need to do here.

No, my mind goes completely blank. There's nothing wrong with my memory - I can tell you how I patched the timed Grand Theft Auto 1 demo over 25 years ago with SoftICE as a kid, but I assume my brain finds no value in tracking bugs.

Re: CheatGPT

#508

Post by brookst »

dinkumthinkum wrote: post #270
brookst wrote: post #29

Earlier quoted context omitted.

I think it just moves the definition of “learning” to a higher level of abstraction: so you know what AI tools to use, how to prompt them, and how to understand their output? I’m reminded of the time when graphing calculators were going to destroy math programs because nobody would “really know” how to do the work. And yet here we are, and math is fine, and calculators are just another tool.

Most math classes are taught without graphing calculators and when they are used it is minimal. This is fundamentally different. This literally is being people trying to substitute actually know or successfully being evaluated to do anything. The advocates of “bringing down higher education with AI” also consider competence to be “ableism”.

I support using AI in education because it will be in the work environment. It seems insane that anyone would want to teach students on slide rules and typewriters and then send them out to a world of computers and word processors.

And I have no idea where your “ableism” comment came from. Just trying to inject some culture war?

Re: CheatGPT

#509

Post by JohnFen »

petilon wrote: post #390
JohnFen wrote: post #317

Earlier quoted context omitted.

> through various kinds of data abstraction, reasoning by analogy, and other techniques similar to what humans do. No, that's exactly not how LLMs work. They are extremely good at predicting what sentences resemble the sentences in their training data and creating those. That's all. People are getting tripped up because they are seeing legitimate intelligence in the output from these systems -- but that intelligence…

I see a lot of comments along the lines of "it's just predicting the next word". But there's evidence that's what humans do as well: "In the last few decades, there has been an increased interest in the role of prediction in language comprehension. The idea that people predict (i.e., context-based pre-activation of upcoming linguistic input) was deemed controversial at first. However, present-day theories of language…

Sure, but that's not what makes humans intelligent.

Re: CheatGPT

#510

Post by JohnFen »

adastra22 wrote: post #358
JohnFen wrote: post #317

Earlier quoted context omitted.

> through various kinds of data abstraction, reasoning by analogy, and other techniques similar to what humans do. No, that's exactly not how LLMs work. They are extremely good at predicting what sentences resemble the sentences in their training data and creating those. That's all. People are getting tripped up because they are seeing legitimate intelligence in the output from these systems -- but that intelligence…

LLM are not fancy Markov chains. These are more than mere statistical prediction. They contain large deeply layered attentional networks which are perfectly capable of representing complex, arbitrarily structured models, trained from the data set or assembled on the fly based on input. I'm sorry, but I think you are about a decade or so out of date in your intuitions for how these things work. (And a decade is a long…

I will grant that my understanding is not complete (and would argue that pretty much everyone else's is incomplete as well), but it's not out of date. I have deliberately avoided forming any opinion about this stuff until I learned more about what the modern approach is. I'm not relying on what I learned a decade ago.
Post reply on HN