Humans Need Not Apply just popped up randomly in my recommendeds. The whole time I was watching it, I was wondering “what would he think of how much ai has advanced lately?” Then I see this video, just posted 2 weeks ago.
At 11:00 minutes… you underestimate the value of the simplicity of the Humans Need Not Apply video. Some parts of it might feel a little dated, self-driving cars haven't quite arrived for example. Still, this video remains, to my mind, the single best introduction to the topic.
I'm not convinced that technology will create mass unemployment. Here in California, the minimum wage for fast food workers was raised to $20/hour. Despite the fears of fast food restaurant closures, there's still a worker shortage. The global population is expected to peak within 40-50 years, and before that decline more and more of the population will be too old to work. This is already happening in Japan and Korea and they NEED automated labor and robots. Also what do I do about this? I lose my job in 20 years due to a robot. So I get a new job.
I couldn’t recall a single url even of articles I have written and published myself yet Myke expects LLMs (which are modeled after how the brain works) to retrieve websites accurately. The point is I personally don’t have to remember article URLs because bookmarks and Google exist. I only need to vaguely remember some keywords to find an article I read years ago and the same is true in principle for AI.
You can be 100% sure that an input to a classical computer program can't "jailbreak" it. These kind of code injection problems are always due doing the wrong thing to the data.
I think this episode helped actualize for me, a lot of the things AI doomers keep talking about. I think it would be very valuable for grey to make an explainer video on some of the ideas he expressed in this video
People individualistically select and ignore what they receive from previous works very selectively as their inspiration. AI remembers the whole work. Not sure how individualistically selective AI can be…
One of the funniest episodes that I can remember. Really enjoyed this one. Also, Myke, what do you mean, maybe you should set up a fake email address just in case? Every last person I know has a whole bunch of those. I thought you of all people would have like a hundred, each for its own particular purpose, with spreadsheets and everything.
14:16 reminds me of my bachelor thesis, on my notebook cover it (for some reason) printed “@2.4GHz frequency” as “H2.4GHz frequency”. On the day of the presentation my professors basically said fvck off do it again . Presentation was great though.
The part of the video that stands out the most to me is that AI/autos are inevitable. We could not stop cars from replacing horses. We could not stop Uber from taking over taxis. We cannot stop ai from taking jobs from humans. We can’t. There’s no point in fighting it, adapt to it. Blockbuster fought against streaming. Netflix embraced it and adapted. Which one will you be?
When Grey references papers, I wish he'd put them in the show notes. (I understand why he doesn't, I'm sure it would be a huge pain in the ass to hunt them down.) I really want to read those AI proofs he mentioned. (Not to mention all the other times over the years when he's referenced papers on various subjects that sound absolutely fascinating.)
I think the car safety thing is not so much that people want their death to be their own fault, but more so that they want it to be _a human's fault._ Which I think probably comes from the same psychology that makes people want their art and music and writing to come from a human instead of an AI: It isn't just about what the art looks like or what the music sounds like or what the writing says, or even what emotions it inspires. It's also about forging a connection with the mind that made it. Because humans like being connected with other humans. And if there's no human mind, there's no human connection. So people want their affairs handled by humans; including death. If you die because another human made a mistake, you can process that, because you're also a human and therefore you could've made a similar mistake in their position. You can relate to that, and understand that, and model that, so you feel a connection. If you die because a computer ran a line of code that didn't fall in your favour, or ran into some kind of glitch, that doesn't resemble anything that could happen in your own brain; and even if it did, there's no consciousness there for you to model. So there's no connection. Besides, people are _used_ to the idea of being killed by people. It's been happening for thousands of years. We have a lot of _practice_ thinking about it and coping with it. Being killed by machines and computers is, relatively speaking, a brand new phenomenon. We hardly have any practice with it.
Right around @43:10 really took a big weight off of my shoulders. I'm not in one of the teams as AI may come for things I do that make money, but I also use AI to do things that make money. It's such a weird position to be in. I really do not love that this is the era of AI, but also, there are some things that I could NOT have done pre-AI-era. Thank you for addressing this.
Making art is extremely easy. I can go outside and take a picture with my phone. That's an exact replica of reality. I could frame it and even try to sell it. Nobody would buy it though, since everybody with a phone can do the same. The difference between making a picture and making art is that you need to create something that people admire. That's not what these "creative" AIs are doing. They're just taking pictures.
Can anyone find the original paper about the claude bot having a mental break down? I think there is one similar in the show notes but id like to see the expirements “transcript” so to speak
@CGP Grey - you should re-visit the "Rules for Rulers" in the context of "What if Ai and robots could do pretty much everything." I.e. - what if the ruler COULD (effectively) do everything himself?
DNA and RNA from viruses are the biological "forget all previous instructions" of the cellular world. A traditional virus is programmed for self replication but the DNA or RNA payload can be anything: malformed proteins, activation or deactivation of a cellular process, programmed apoptosis, etc.
As I see more AI offerings in more places, particularly in text-based models, I always ask myself "who thinks I need more of this in my life?". If I ask it anything important to me or about anything I know about, it constantly lies. At least with my uncle Richard, he lied to me as a kid with a wink and a nod. I don't need AI to lie to me, and if there's even a possibility of that, I have absolutely no use for it. Also, general purpose AI has literally been "five years away" for about 60 years, which is to say the most useful thing we've learned in the pursuit of AI is that we don't actually know what we're expecting it to do nor are our theoretical models of consciousness useful - either because we can't replicate them, or when we do, they come up short.
As a normal person, I have no idea what they are talking about with AI being political. And they talked around it in such vagueness that even after listening to it I am still confused.
Right? The lack of detail was even a bit awkward, when they really got concrete about the analogies. Can we hear *less* about the analogies and more about the real thing?
They both explained themselves fairly clearly there. They dont think AI is 'political' in some sense that the right is pro and the left is anti or something, but 'political' in the way that the conversations are discussed. To remove the term from the explanation, they find that people have very solidly held beliefs on the topic, and that therefore the discussion around AI is less organic than it was before people decided on a 'party line' to hold.
I based some part of initial reflexion of my master thesis in 2018 on this video (What is the impact of automation process in the supermarket environement?). Since then, I worked on project management for automation implementation. 10 years later, here we are.
I find the collated intelligence more important than the reasoning engine side, but we are playing with fire how much changing our decisions can affect the world. This is definitely a pandemic situation.
They have well paid engineers writing psychological guiding to guide better output generation. The more specific you are, the better output you will get. If not, it is all dependent on context and randomness.
It's not fair use. It's not constructive or educational. It is copyright infringement, but they are trying to pass it off, in just the opposite way of Disney claiming copyright is forever
I think the innovation is not much faster, I think things are changing less, but people are really investing in machine learning technology and developing it commercially. But e are definitely on the sigmoid curve. GPT-4 has gotten a lot better "creative" and "useful" now that the "omni" models have released. But I think a lot of people are also becoming more annoyed with the proliferation? And everyone is becoming more aware of its shortcomings.
I was using the GPT-3 text models before ChatGPT. Even GPT-2 seemed crazy, even though it didn't feel useful. ChatGPT blasted through the ceiling and past the moon. And they most likely did that to gather a lot of data to make it more like humans.
It's a bit different. In planes, the automation just reduces the workload, the pilots have to monitor everything, they control where and how it flies, just means not manually moving it all the time. It is very human and requires them at most stages. Most importantly, the planes are tested to try and keep them in perfect working order.
As someone who has tried out a recent ish version of Tesla's full self driving and works on the roads I would argue that self driving cars are already much better than the average driver. FYI I hold the positions that the average driver shouldn't be allowed behind the wheel, and Tesla's full self driving is extremely flawed especially when navigating parking lots.
Hey guys. I’m checking in as a “this is just a calculator” guy (although this somewhat oversimplifies the position). LLM train themselves on the data that’s input into them. In the cash of the popular LLM (chat gpt, bard, etc) that data is the everyday conversations of people on the internet. Given that, it makes perfect sense that it comes off person like. If you took a calculator and programmed it to say “I am alive and suffering” it would give you that message, even though it doesn’t know what that means. Ditto with LLM. We input data that implies sentience (since it came from sentient people) so the outputs we get mirror that. That doesn’t mean it’s actually alive. Imagine if I was creating a new operating system, but instead of hiring engineers I crowdsourced the coding of the operating system, and put all code submitted into the software without checking it. Imagine all the goofy things that OS would do. I don’t see how this is different.
I agree with Myke about companies profiting from 'fair use' and not willing to jump onto this bandwagon. The quality of the outputs is really low, and I realise that if you use LLMs for little stuff, it will coax you into trusting it for bigger stuff. And the hallucinations you speak of: so true. And look at humans: they rumour and hallucinate collectively often: we still can't get the autism/vaccine linkage broken--that mind virus persists. We already disagree on what reality is with our media bubbles; AI will supercharge that. Look at our politics: People think immigrants take their jobs and then go to ChatGPT to prove their case. CHATGPT is the threat to their job and capitalism is always looking for ways to make it so that Humans Need Not Apply. AI is a particularly pernicious form of dystopian capitalism; The movie Idiocracy dives into it clearly. And it seems to me that dropping this huge error of computer sentience is the best path forward. I feel that once AI starts making music and art that is more popular than human production... there will just be no point in being human anymore nor being on the internet. Because it is about trust. And we can't trust this stuff. If you're unnerved, spooked, and otherwise feeling bad about this, it seems to me that we should perhaps focus on bettering society for humans instead of dumping all this effort in service of machines. "these tools are not like anything else" Let that ring out. Please. 10 years from now, I'm going to catch a lot of flak from this comment. And I hope I will be able to survive in the dystopia that may come.
The most advanced language models out there will insist that the word "strawberry" only contains two instances of the letter R. Even after you literally spell it out for them. Even after getting them to agree that there are in fact three, if you ask them they'll reply that there are two. I live in the English-speaking world, so I can't even imagine the weirdness that would come from models trained on websites using other languages and written scripts...
The observation grey makes towards the end is subtle and profound. That not just constructs but ALL our ideas are experiencing (subject to?) selective pressure to evolve is such a key idea. It has caused my brain to start re-calibrating everything into a new frame of reference. Well played Sir!
Mike, no, not everyone is more vulnerable on an airplane. I've done it enough, and I believe in its capabilities enough, that it usually causes me zero extra stress to be on an airplane. Being on the airplane is often one of the least stressful parts of travel. (Caveat: this is assuming I don't need to sleep. Sleeping on an airplane is generally pretty awful, lol)
Can someone please tell me where to find what they were saying about Claude talking to itself and having a mental breakdown? I Have checked the links and i can't find it.
