Live data from Hacker News

How accurate have Ed Zitron's AI skeptic predictions been?

danluu.com

701–710 of 1001 posts

Re: How accurate have Ed Zitron's AI skeptic predictions been?

#701

Earlier quoted context omitted.

I think the three valid criticisms you listed are valid, but there are other quite a few other IMO-valid criticisms of AI. Here are a few as I see them: - AI centralizes power in the hands of capital, rendering those who can afford compute hardware vastly more capable than those who cannot and thus increasing social stratification - AI is generally trained on the creative output of humanity without those who train it…

> - AI centralizes power in the hands of capital, rendering those who can afford compute hardware vastly more capable than those who cannot and thus increasing social stratification Looks to me like I can run Claude Code without being able to afford my own datacenter. > - AI produces large amounts of mid-tier content, making it harder to discover exceptional human-created creative content produced after AI started to…

> Looks to me like I can run Claude Code without being able to afford my own datacenter.

Well, unless of course you want to train your own LLM, or do some biochemistry (and increasingly just regular health stuff) or cybersecurity. These capabilities are not made available for plebs like you or I.

Re: How accurate have Ed Zitron's AI skeptic predictions been?

#702
Accurate or not, there are many risks other than those that could bring AI companies value to 0. E.g., systems not requiring passing over all the data continuously, but in small regions running at L2-L3 cache memory speed. Once someone figures out that, the need for high bandwidth memory would end.

Re: How accurate have Ed Zitron's AI skeptic predictions been?

#703

One thing I’m observing in these comments is a willingness of folks to project their own predictions onto Ed’s statements when validating their plausibility. Eg. “I think he’s wrong about the timing but I do expect AI companies to go to zero.” You can do that, but then you’re no longer discussing his predictions. You’re discussing your predictions, and your own positioning. Those differ from Dan’s essay, which engage…

This is an article that only cites Zitron’s opinions about subjective model quality. You’re tisk-tisking people to be objective about a bunch of words saying that the author’s opinions are better opinions than the guy he’s talking about OP

Some guy wrote that he’s grumpy because he couldn’t sleep and decided to dunk on an internet personality he doesn’t like, it’s not the ceremonial placement of the ur-kilogram

Re: How accurate have Ed Zitron's AI skeptic predictions been?

#704
post #683
post #376

Earlier quoted context omitted.

Be kind. Don't be snarky. Converse curiously; don't cross-examine. Edit out swipes. Comments should get more thoughtful and substantive, not less, as a topic gets more divisive. When disagreeing, please reply to the argument instead of calling names. "That is idiotic; 1 + 1 is 2, not 3" can be shortened to "1 + 1 is 2, not 3." Please don't fulminate. Please don't sneer... https://news.ycombinator.com/newsguidelines.h…

It's my fault. I'll follow the rules more carefully.

Many thanks!

Re: How accurate have Ed Zitron's AI skeptic predictions been?

#705

This kind of back and forth reminds me of the Dot Com bubble circa 1999. There were endless articles saying that the internet was a new golden era for mankind and that the hyperbolic company valuations were justified; the detractors cried bullocks. I use AI every day as I assume everyone else on Hacker News does. Will the S&P 500 drop 20%? I have no idea. If you know, please let me know so I can adjust by 401K.

> assume everyone else on Hacker News does

Eh, not to be rude/crass... but that would be inaccurate, if not hopeful. Myself and others don't, despite mandates; in fact, I aim to see this 'left behind' promise we heard years ago. Short of writing the at-will employment paperwork for HR myself, I'm not seeing it.

Escalations continue to fill my days. Turns out, people are somewhat correct: results matter. I'd say moreso than the tools we 'choose' (or skip, in this case). My null on the token scoreboard remains unnoticed/inconsequential, the work I've done has not.

All to raise a bit of timeless advice from Wu-Tang: diversify.

Re: How accurate have Ed Zitron's AI skeptic predictions been?

#706
post #158

Earlier quoted context omitted.

This is exactly right. Now, that does mean something about Zitron: he’s a pundit, not an expert or a forecasting genius. You could point to any number of analogous booster types. Twitter/X somehow loves pushing these people onto my recommended feed. I remember reading breathless threads about how o3 was going to single-handedly end white collar work. There are just more of that sort of person, so none of them in part…

In my experience, people who make fun of AI's ability to eliminate white-collar work are exactly the same people who are white-knuckling it, hoping they can make another $300k next year to buy that Rivian. People who don't need a salary look at the situation more objectively and, in general, can see that a lot of white-collar work is in peril.

IME many of the people who truly believe it's replacing white collar workers are being replaced are invested in the stock market bubble.

Re: How accurate have Ed Zitron's AI skeptic predictions been?

#707

One thing I’m observing in these comments is a willingness of folks to project their own predictions onto Ed’s statements when validating their plausibility. Eg. “I think he’s wrong about the timing but I do expect AI companies to go to zero.” You can do that, but then you’re no longer discussing his predictions. You’re discussing your predictions, and your own positioning. Those differ from Dan’s essay, which engage…

Why is Zitron’s repute evaluated entirely on the basis of failed predictions? Predictions are incredibly hard. AI enthusiasts and thought leaders have made so many demonstrably incorrect predictions it’s hard to keep track. Based on this metric, Altman and Amodei should never be taken seriously again.

I think it's fair to say predictions that Altman and Amodei make should never be taken at face value, as well as Zitron. That's fine. But that doesn't have any bearing on Dan Luu's claims. This feels like an example of what he talks about in the article when saying that people respond to his claims by pointing at something entirely different. That is to say, whether AI enthusiasts make silly predictions doesn't mean that these companies aren't going to be profitable, or that any of Zitron's predictions are any good either.

Re: How accurate have Ed Zitron's AI skeptic predictions been?

#708

Earlier quoted context omitted.

> The prediction was early, overestimated how long of a tail RIM would experience, but how wrong was it? Was the prediction of some utility? Wrong enough that the utility is seriously diminished. Predicting a specific quantity dropping to a specific level at a specific date is a lot more valuable than saying “those guys are cooked”. And even if some minuscule utility existed: why should predictors be so coddled by th…

RIM is dying now -> useful for stock trading RIM is dying in the next four years -> useful for life planning. Anyway, i just wanna get on record that i predict an ai bubble pop event in the next 12 months.

> Anyway, i just wanna get on record that i predict an ai bubble pop event in the next 12 months.

If it happens on Sep 2nd 2027 you'd be wrong though.

Re: How accurate have Ed Zitron's AI skeptic predictions been?

#709

Danluu claims Ray Kurzweil's predictions are wrong. I asked > list 50 predictions by Ray Kurzweil that he made before 2005, the date he made the prediction, and whether to not his prediction was correct. If it was correct, give some proof. Of the 50 that were listed, 43 were correct, 7 incorrect.

You "asked"? Who (or more likely, what) did you ask? Dan Luu's analysis of Kurzweil is in this article: https://danluu.com/futurist-predictions/ In the "Appendix: detailed information on predictions" he lists the predictions, and their truth or lack thereof. He was using a list on Wikipedia: https://web.archive.org/web/20170225013846/https://en.wikipe... For example Kurzweil apparently predicted in 2019 that: > Blind…

Maybe we have different interpretations of whether or not a prediction is correct or not.

If someone predicts in 1999 predicts self driving cars by 2020 and it happens in 2030. That seems different than predicting something we see no indication of ever happening. Like if they predicted 2020 and it happened in 2021 is that a fail. To me, no. The prediction is that the tech will exist at or around that date, not that the date is exact. Claiming it's a fail for being 1 second late is not a rational position and not at all in the spirit the predictions were made.

There's also issues like being directionally correct. Example: Someone says computers will be in everything. You can say "there's no computer's in fruit!, FAIL!" or you can look at the explosion of IoT devices and decide it was mostly right?

I get 17% correct if you're absolutely strict, 64% correct if you're charitable

2009

* Most books will be read on screens rather than paper.

The charitable interpretation is that most reading happens on screens, not paper. This is true today.

* Most text will be created using speech recognition technology.

False

* Intelligent roads and driverless cars will be in use, mostly on highways.

False in 2009, False in 2026 but directionally true. If you live in an area with Waymo you see them all time. I've driven down Olympic Blvd in Los Angeles and had my car surrounded by 5 Waymo cars at once. So is this false because it didn't happen by 2009 or is at least directionally true because it's happening, we see evidence of it happening, vs if we saw zero evidence then we could 100% say it's false.

* People use personal computers the size of rings, pins, credit cards and books.

rings, pins and credit cards, no, books, true. Smartphones are smaller than books. Maybe you could make the argument those are not personal computers. I think that is debatable. Even then, you can by PIs or Mini-PCs that are book size.

* Personal worn computers provide monitoring of body functions, automated identity and directions for navigation.

Arguably true. phones provide directions for navigation and are worn in pockets. Fitbits came out only a few years later. Id is not automated though.

* Cables are disappearing. Computer peripherals use wireless communication.

Arguably true. most laptops, all phones, most mice, keyboards, joypads, etc. are all wireless.

* People can talk to their computer to give commands.

False/True. Was possible was not common. That said, Siri shipped in 2011 so 2 years off.

* Computer displays built into eyeglasses for augmented reality are used.

False if you mean mainstream.

* Computers can recognize their owner's face from a picture or video.

Face ID shipped in 2017. Is that to far off?

* Three-dimensional chips are commonly used.

I'm not sure what this means.

* Sound producing speakers are being replaced with very small chip-based devices that can place high resolution sound anywhere in three-dimensional space.

False,

* A $1,000 computer can perform a trillion calculations per second.

True, happened in 2008 with the ATI Radeon HD 4850

* There is increasing interest in massively parallel neural nets, genetic algorithms and other forms of "chaotic" or complexity theory computing.

Happened in 2012 so 3 years off

* Research has been initiated on reverse engineering the brain through both destructive and non-invasive scans.

Based on the actual words, this is true and was true before the prediction.

* Autonomous nano-engineered machines have been demonstrated and include their own computational controls.

False

...continued...

Re: How accurate have Ed Zitron's AI skeptic predictions been?

#710

Earlier quoted context omitted.

You "asked"? Who (or more likely, what) did you ask? Dan Luu's analysis of Kurzweil is in this article: https://danluu.com/futurist-predictions/ In the "Appendix: detailed information on predictions" he lists the predictions, and their truth or lack thereof. He was using a list on Wikipedia: https://web.archive.org/web/20170225013846/https://en.wikipe... For example Kurzweil apparently predicted in 2019 that: > Blind…

Maybe we have different interpretations of whether or not a prediction is correct or not. If someone predicts in 1999 predicts self driving cars by 2020 and it happens in 2030. That seems different than predicting something we see no indication of ever happening. Like if they predicted 2020 and it happened in 2021 is that a fail. To me, no. The prediction is that the tech will exist at or around that date, not that t…

...continued from above (part 2) ...

2019

* The computational capacity of a $4,000 computing device (in 1999 dollars) is approximately equal to the computational capability of the human brain (20 quadrillion calculations per second).

False (though if we're taking FP4 it's only 1 order of magnitude off9

* The summed computational powers of all computers is comparable to the total brainpower of the human race.

False

* Computers are embedded everywhere in the environment (inside of furniture, jewelry, walls, clothing, etc.).

The charitable interpretation is this true. There are plenty of all of those things. Jewelry (apple watch, Oura ring, Walls = LED lighting systems, furniture = message chairs with apps)

* People experience 3-D virtual reality through glasses and contact lenses that beam images directly to their retinas (retinal display). Coupled with an auditory source (headphones), users can remotely communicate with other people and access the Internet.

These special glasses and contact lenses can deliver "augmented reality" and "virtual reality" in three different ways. First, they can project "heads-up-displays" (HUDs) across the user's field of vision, superimposing images that stay in place in the environment regardless of the user's perspective or orientation. Second, virtual objects or people could be rendered in fixed locations by the glasses, so when the user's eyes look elsewhere, the objects appear to stay in their places. Third, the devices could block out the "real" world entirely and fully immerse the user in a virtual reality environment.

False, though all of that has been demonstrated

* People communicate with their computers via two-way speech and gestures instead of with keyboards. Furthermore, most of this interaction occurs through computerized assistants with different personalities that the user can select or customize. Dealing with computers thus becomes more and more like dealing with a human being.

Charitable version is Siri, Alexa. And arguably it's clear it will happen with LLMs so not far off.

* Most business transactions or information inquiries involve dealing with a simulated person.

False in 2019 but seems directionally true in 2026. So many businesses use AI chat and or AI customer service. Even the DMV is now AI.

* Most people own more than one PC, though the concept of what a "computer" is has changed considerably: Computers are no longer limited in design to laptops or CPUs contained in a large box connected to a monitor. Instead, devices with computer capabilities come in all sorts of unexpected shapes and sizes.

True? Most people own a phone and a smart TV or a phone and tablet, or a phone and watch or a phone and video game system.

* Cables connecting computers and peripherals have almost completely disappeared.

Arguably false, otherwise I wouldn't have so many cables.

* Rotating computer hard drives are no longer used.

Directionally true. The average person has a phone, tablet, PC, TV, PS5, Switch, with SSD, not rotating HD. Hard drives are still common in data centers and geek media hubs

* Three-dimensional nanotube lattices are the dominant computing substrate.

False

* Massively parallel neural nets and genetic algorithms are in wide use.

True in 2026, No idea if it was true behind the scenes in 2019.

* Destructive scans of the brain and noninvasive brain scans have allowed scientists to understand the brain much better. The algorithms that allow the relatively small genetic code of the brain to construct a much more complex organ are being transferred into computer neural nets.

No idea

* Pinhead-sized cameras are everywhere.

False, but if you want to be charitable, cameras everywhere (Amazon Ring, Google Nest, etc...) are everywhere.

* Nanotechnology is more capable and is in use for specialized applications, yet it has not yet made it into the mainstream. "Nanoengineered machines" begin to be used in manufacturing.

No idea. It's true it's not yet made it into the mainstream. But there are many "nano-materials"

* Thin, lightweight, handheld displays with very high resolutions are the preferred means for viewing documents. The aforementioned computer eyeglasses and contact lenses are also used for this same purpose, and all download the information wirelessly.

Arguably true. The majority of phones have a very high resolution display and it's where most data is viewed. The 2nd part is false.

* Computers have made paper books and documents almost completely obsolete.

Again, charitably, the majority of text and documents are not digital.

* Most learning is accomplished through intelligent, adaptive courseware presented by computer-simulated teachers. In the learning process, human adults fill the counselor and mentor roles instead of being academic instructors. These assistants are often not physically present, and help students remotely.

False, maybe directionally true. I know lots of people and kids that LLMs, not humans to learn things.

* Students still learn together and socialize, though this is often done remotely via computers.

True. Tons of children socialize remotely. They also learned remotely (COVID)

* All students have access to computers.

True? Does this fail on the word "all" or does it pass because it's mostly true.

* Most human workers spend the majority of their time acquiring new skills and knowledge.

False

Post reply on HN