Live data from Hacker News

GPTs Are GPTs: An Early Look at the Labor Market Impact Potential of LLMs

arxiv.org

201–210 of 235 posts

Re: GPTs Are GPTs: An Early Look at the Labor Market Impact Potential of LLMs

#201

With little kids I am increasingly conscious about how to best equip them for the world when they are adults. I think previous playbook will need to be redrafted. 'Resilience' and 'Critical Thinking' are two things I am thinking are key any others?

It's often said that critical thinking is the cornerstone of a well-rounded education. Universities and colleges around the world extol the virtues of critical thinking and its ability to elevate our intellectual prowess. However, I'd like to argue that critical thinking, or more specifically, the way it's often employed in certain academic contexts, has become overrated and counterproductive to meaningful, constructive discourse.

In American college policy debates, a common phenomenon has emerged where "kritiks" (critiques) and "critical theorists" wield their tools of analysis in a manner that stifles the flow of ideas and inhibits the development of practical, solutions-oriented discussions. For those unfamiliar with policy debate, it is a competitive speaking activity where teams of two advocate for and against a resolution that typically proposes a policy change.

Kritiks are arguments that challenge the assumptions, methodology, or discourse used by the opposing team. While they can serve a valuable purpose in exposing biases, promoting introspection, and advocating for marginalized perspectives, they have also evolved into a means of derailing debates by focusing on abstract philosophical points rather than addressing the policy issue at hand.

Critical theorists, heavily influenced by postmodernist and marxist philosophy, have taken what was originally a quirk of Policy Debate and brought it to the mainstream. This has ended up with serial policy failure among the left wing, and has decimated our social science academic community.

"Critical Thinking" is overrated. We need constructive thinking, like what you get in engineering school.

Re: GPTs Are GPTs: An Early Look at the Labor Market Impact Potential of LLMs

#202
post #170

Earlier quoted context omitted.

If you want to continue with semantics, the worker did not expect to hit his thumb until the hammer bounced off of it, hence the tool did not work as expected. Until regular hammers are able to put nails into wood by themselves, we are talking about the success of the complete system that includes both the hammer and its user. Either way, when working with humans we already deal with plenty of misses and mistakes. Pr…

Nice packpedal, except now what you said makes no sense anyway because the discussion was about tools doing what you expect them to do, not "tool + human systems" doing what you expect them to. These are two different things. Do you think a carpenter who hits themselves with a hammer blames the hammer or themselves? Or are you going to unironically tell me they would blame the hammer + human system?

>not "tool + human systems" doing what you expect them to. These are two different things.

Please explain how?

>Or are you going to unironically tell me they would blame the hammer + human system?

As your tooling gets more complex, yes it is very easy to have a non-zero blame assignment to each party. Look at any human+machine system where complex failure conditions can occur.

Re: GPTs Are GPTs: An Early Look at the Labor Market Impact Potential of LLMs

#203

Earlier quoted context omitted.

Not sure if you tried GPT-4 but my experience with 4 is quite different. It has been quite bullshit free, though not completely. For example I asked it to contrast oral and injectable semaglutide formulations. It did a bang up job. One thing I always do is ask it for evidence. And then I look the references up. Sometimes the references don’t say exactly what it said they will. I come back and have a discussion with i…

> One thing I always do is ask it for evidence. And then I look the references up. Sometimes the references don’t say exactly what it said they will. Can you give an example? I'm surprised there's room in there for actual resolvable URLs or paper references.

If you’re discussing a scientific topic just ask it for references, it generally gives the first author, year and journal along with the title, for most reasonably cited papers!

Re: GPTs Are GPTs: An Early Look at the Labor Market Impact Potential of LLMs

#204

Earlier quoted context omitted.

ChatGPT is nothing more than long-form autocomplete. The specifics of approach are irrelevant.

A human brain is nothing more than long-form autocomplete. The specifics of approach are irrelevant.

Why do you think that?

Re: GPTs Are GPTs: An Early Look at the Labor Market Impact Potential of LLMs

#206

I asked Chat GPT which antacid medications are contraindicated for some medication I'm on. Easily found through NICE. It made up a severe risk of death taking a very common medicine combo. It was super convincing, even giving information on how long to avoid taking them together. It was pure bullshit. I think as much as hyping the benefits we need to hype the flaws and dangers. If the public at large learn to trust t…

> It made up a severe risk of death taking a very common medicine combo.

Probably trained on too many of those SEO hyper optimized medical sites people get when they search symptoms and such.

Re: GPTs Are GPTs: An Early Look at the Labor Market Impact Potential of LLMs

#207
post #6
post #2

==Occupations with no labeled exposed tasks== Agricultural Equipment Operators Athletes and Sports Competitors Automotive Glass Installers and Repairers Bus and Truck Mechanics and Diesel Engine Specialists Cement Masons and Concrete Finishers Cooks, Short Order Cutters and Trimmers, Hand Derrick Operators, Oil and Gas Dining Room and Cafeteria Attendants and Bartender Helpers Dishwashers Dredge Operators Electrical…

Unfortunately, the great majority of these jobs are already being automated with basic robotics

Definitely

Re: GPTs Are GPTs: An Early Look at the Labor Market Impact Potential of LLMs

#208

I asked Chat GPT which antacid medications are contraindicated for some medication I'm on. Easily found through NICE. It made up a severe risk of death taking a very common medicine combo. It was super convincing, even giving information on how long to avoid taking them together. It was pure bullshit. I think as much as hyping the benefits we need to hype the flaws and dangers. If the public at large learn to trust t…

I asked it "What is the accuracy and confidence level (you must rate it out of 10) that the above answer is correct ?" It always generates a random number when it is caught making up bullshit sometimes claims that the information 10/10 accurate and LLM's always provides accurate information.

OpenAI did exact the test for GPT-4. The raw, non-fine-tune GPT-4 is quite good at predicting confidence level ("highly calibrated" by their words). But the RLHF fine-tuning process seems ruin its calibration. Figure 8 on page 12 of GPT-4 Technical Report shows this dramatic changes before & after fine-tuning.

Re: GPTs Are GPTs: An Early Look at the Labor Market Impact Potential of LLMs

#209
post #31
post #13

Earlier quoted context omitted.

Programming skills that do not need critical thinking are more susceptible of being influenced indeed. Don't forget that a lot of science requires computer programming these days. This is the root of it: The more "genericc" your work is. The more its "out there on the internet" the more GPT can learn about it.. So, a lot of engineers that are just doign teh same old trick: Writing HTTP endpoints, parsing json. Mappin…

> So, a lot of engineers that are just doign teh same old trick: Writing HTTP endpoints, parsing json. Mapping data types.. Yes that could be automated. And to be fair, automation for all of that already pretty much exists.

It is exists where it can be supported. Lots of small businesses don't have the bandwidth to maintain additional infra that automates this sort of work.

Which sorta brings me back around: it's likely the Big Corps that are going to be trialing GPT first because they have the excess money and resources to play with it. How useful will it be in the end?

Re: GPTs Are GPTs: An Early Look at the Labor Market Impact Potential of LLMs

#210

Earlier quoted context omitted.

I've been using Bing AI instead of Google for a few days to test it out. It gives responses with citations. A few times already it has hallucinated details that aren't on the cited website at all, but that do sound very plausible (for example something about the horns of a bull pointing up when I asked about the etymology of "bullish".) I think it comes down to the fact that most stuff is actually like 90% bullshit s…

I've said this already today on another comment, but the manual from OpenAI themselves pretty much says the thing is not to be trusted for pretty much anything.

Same for Bing chat.
Post reply on HN