Hype Critic Compares AI to the Roomba

Today on the AI Daily Brief, we're reading an essay arguing that AI is.

Overhyped and will be nothing more significant.

Than Roombas and then talking about why.

It'S incredibly, incredibly wrong.

The AI Daily Brief is a daily podcast and video about the most important news and discussions in AI. To join the conversation, go to the Discord linked in our show Notes. Hello friends. There is a very particular emergent genre.

Of essays which I believe will be.

Endlessly ridiculed in the future. They are basically all of the argument that the AI hype is ridiculous and that it's really no more powerful than any other technology, and that it's really just the manufacturer of big tech, and that we won't really care about all this and it will all seem stupid in a few years. This is not the first time that this genre of essay has made waves. Perhaps the most famous example of this was Paul Krugman's essay from 1998 that gave us this famous quote. The growth of the Internet will slow drastically as the flaw in Metcalfe's Law, which states that the number of potential connections in a network is proportional to the square of the number of participants, becomes apparent. Most people have nothing to say to each other. By 2005 or so, it will become clear that the Internet's impact on the economy has been no greater than the fax machines.

Mm, nailed that one, Krugman.

So today we are reading, or rather my AI is reading, a piece called Press Pause on the Silicon Valley Hype Machine. It's by Julia Anguin and was published in the New York Times. I don't think the source of its publishing is an accident, but let's read this and then we'll discuss it. Or rather, let's let an AI version.

Of me read this and then a real version of me will discuss it. It's a little hard to believe that just over a year ago a group of leading researchers asked for a six month pause in the development of larger systems of artificial intelligence, fearing that the systems would become too powerful. Should we risk loss of control of our civilization? They asked Their was no pause. But now, a year later, the question isn't really whether AI is too smart and will take over the world. It's whether AI is too stupid and unreliable to be useful. Consider this week's announcement from OpenAI's chief executive, Sam Altman, who promised he would unveil new stuff that feels like magic to me, but it was just a rather routine Update that makes ChatGPT cheaper and faster. It feels like another sign that AI is not even close to living up to its hype. In my eyes, it's looking less like an all powerful being and more like a bad intern who's whose work is so unreliable that it's often easier to do the task yourself. That realization has real implications for the way we, our employers, and our government should deal with Silicon Valley's latest dazzling new new thing. Acknowledging AI's flaws could help us invest our resources more efficiently and also allow us to turn our attention toward more realistic solutions. Others voice similar concerns. I find my feelings about AI are actually pretty similar to my feelings about blockchains. They do a poor job of much of what people try to do with them. They can't do the things their creators claim they one day might, and many of the things they are well suited to do may not be altogether that beneficial, wrote Molly White, a cryptocurrency researcher and critic, in her newsletter last month. Let's look at the research. In the past 10 years, AI has conquered many tasks that were previously unimaginable, such as successfully identifying images, writing complete, coherent sentences, and transcribing audio. AI enabled a singer who had lost his voice to release a new song using AI trained with clips from his old songs. But some of AI's greatest accomplishments seem inflated. Some of you may remember that The AI model ChatGPT4 aced the uniform bar exam a year ago. Turns out that it scored in the 48th percentile, not the 90th, as claimed by OpenAI, according to a reexamination by the MIT researcher Eric Martinez. Or what about Google's claim that it used AI to discover more than 2 million new chemical compounds? A reexamination by experimental materials chemists at the University of California, Santa Barbara, found scant evidence for compounds that fulfill the trifecta of novelty, credibility, and utility. Meanwhile, researchers in many fields have found that AI often struggles to answer even simple questions, whether about the law, medicine, or voter information. Researchers have even found that AI does not always improve the quality of computer programming, the task it is supposed to excel at. I don't think we're in cryptocurrency territory, where the hype turned out to be a cover story for a number of illegal schemes that landed a few big names names in prison. But it's also pretty clear that we're a long way from Mr. Altman's promise that AI will become the most powerful technology humanity has yet invented. Take Devin, a recently released AI software engineer that was breathlessly touted by the tech press a flesh and Bones software developer named Carl Brown decided to take on Devon, a task that took the generative AI powered agent over six hours. Took Mr. Brown just 36 minutes. Devon also executed poorly, running a slower, outdated programming language through a complicated process. Right now, the state of the art of generative AI is it just does a bad, complicated, convoluted job that just makes more work for everyone else, Mr. Brown concluded in his YouTube video, Cognition. Devin's maker responded by acknowledging that Devon did not complete the output requested, and added that it was eager for more feedback so it can keep improving its product. Of course, AI companies are always promising that an actually useful version of their technology is just around the corner. GPT4 is the dumbest model any of you will ever have to use again, by a lot of, Mr. Altman said recently while talking up GPT5 at a recent event at Stanford University. The reality is that AI models can often prepare a decent first draft. But I find that when I use AI, I have to spend almost as much time correcting and revising its output as it would have taken me to do the work myself. And consider for a moment the possibility that perhaps AI isn't going to get that much better anytime soon after. All the AI companies are running out of new data on which to train their models, and they are running out of energy to fuel their power hungry AI machines. Meanwhile, authors and news organizations including the New York Times, are contesting the legality of having their data ingested into the AI models without their consent, which could end up forcing quality data to be withdrawn from the models. Given these constraints, it seems just as likely to me that generative AI could end up like the Roomba, the mediocre vacuum robot that does a passable job when you are home alone, but not if you are expecting guests. Companies that can get by with Roomba quality work will of course still try to replace workers. But in workplaces where quality matters and where workforces such as screenwriters and nurses are unionized, AI may not make significant inroads. And if the AI models are relegated to producing mediocre work, they may have to compete on price rather than quality, which is never good for profit margins. In that scenario, skeptics such as Jeremy Grantham, an investor known for correctly predicting market crashes, could be right that the AI investment bubble is very likely to deflate soon. The biggest question raised by a future populated by unexceptional AI, however, is existential. Should we as a society be investing tens of billions of dollars our precious electricity that could be used toward moving away from fossil fuels and a generation of the brightest math and science minds on incremental improvements in mediocre email writing. We can't abandon work on improving AI. The technology, however middling, is here to stay and people are going to use it. But we should reckon with the possibility that we are investing in an ideal future that may not materialize. Hello friends.

Quick note before we get back to the show. I'm so excited to share that Superintelligent is now live. Superintelligent is a platform for fast, fun and super practical useful AI learning.

We have something like 300 video tutorials adding 30 to 50 each week covering.

Every topic in AI you can imagine, from LLMs to image generators to case studies use cases. Basically everything that tells you how to use AI and what to use it on. In short, fast four to seven minute.

Tutorial videos which are paired with step.

By step instructions that help you actually.

Use these tools as well.

It's $20 a month for unlimited access and I would love to see you there.

Check it out at Be Super AI. That's BeSuper AI.

Alright, so back to NLW here, the.

Real non robot nlw.

Like I said, this is very clearly a part of this genre of essay that's trying to say that this technology in particular AI is overhyped and not.

Really going to go anywhere.

So let's talk about what's wrong with this particular argument. Some of it is just clear logical fallacies, particularly the false analogy fallacy with crypto. In other words, because crypto was overhyped.

This must be two 1 these two.

Things are completely unrelated. The fact that they involve computers is not a sufficient condition for one to reflect the other. And two this one reads even less convincing in a week when by a 2 to 1 margin the house just voted for pro crypto regulations and the chief antagonist of crypto in the us, the SEC was just handed a crushing political defeat by which they had to approve an Ethereum spot ETF over their strenuous concerns because it was not politically.

Tenable to do otherwise.

Anyways, I digress. The bigger point is the false analogy fallacy. I also think it's funny that this new generation of careerist AI critics is trying to call back to and reference the careerist crypto critics who've just been.

Proven wrong and wrong over and over again.

Next problem with this essay is the cherry picking of examples of specific companies over promising specific companies. Overpromising does not necessarily reflect an entire industry or an entire category of technology not actually working. It would be a weak argument for me to cherry pick a bunch of.

Things that are working, and it's an.

Even weaker argument to just pick some where companies overpromised. The biggest issue is how much of this is just raw opinion. Now, of course, this is an opinion column, so I can't really hate on her too much for that, but most of the argument here comes down to I don't know.

It doesn't seem all that good to me.

Perhaps the most absurd argument is that.

AI isn't going to get that much better.

If you have any sense or questions about that, take one look at the comparison between video generation a year ago.

And today, and then come back and talk to me. I'm being rhetorical here.

If you're listening to this, I'm pretty sure you're on this page as well. Like I said, the reality is that.

This is a genre.

It's written by a category of opinion writer who make their lives around the public's mistrust of technology. A mistrust of technology which, by the way, technologies bear a lot of complicity in, but which has at this point certainly gotten to almost fetish levels. This genre is published by news outlets who are looking for an alternative narrative to AI will change everything. And interestingly, perhaps ironically, even in part, it is those it will change everything pieces that have made many folks unable to see the change that's happening right in front of them. People have spent so much time reading about how entire categories of jobs are going to be gone tomorrow that when that doesn't happen, they think AI must be bs. Because of that, they're completely missing what's actually happening. A recent LinkedIn and Microsoft study found that 75% of knowledge workers are using.

AI at work, but even more remarkably.

78% of them are doing so outside of the purview of their bosses in what Microsoft and LinkedIn call BYO AI.

Or bring your own AI.

Basically, you have more than 2/3 of knowledge workers effectively smuggling AI into the office because they find it so valuable that they don't want to be told they're not allowed to do that anymore. How one can look at that and.

Believe that this isn't going to be.

Transformative is very hard for me to understand. Again, these aren't people who are being told by their boss that they have to use this technology. These are people who are hiding from their bosses that they're using this technology because they don't want to be told they're not allowed to. Lastly, on this point, that progress is just going to stall.

First of all, I think it's absurd.

But even if it wasn't, and even if it did stall, the power in these systems today would already be completely transformative in ways that would take years to solve. The reality is, no matter how many critics get featured in op ed pieces between now and then, the AI economy is here. It's getting bigger every day, and it's going to eat everything. Anyways, friends, that will do it. Little bit of an exploration for your Memorial Day weekend. Appreciate you listening or watching as always. And until next time, peace.