Comment by EGreg | Hacker News Reader

Comment by EGreg | Hacker News Reader

Imagine if Microsoft invested $8B in Bitcoin instead of in OpenAI.

Their stock would be worth a lot more today. That’s just a fact at this point, by the numbers.

Now they have to mark down their speculative investment. But of course OpenAI was way more on-brand for MS, and they had to lead the hype, being the kind of company they were, at the time it made sense from an optics point of view.

I have been comparing the AI hype bubble to the Web3 hype bubble since the beginning, but most of HN likes AI far more and doesn’t want to accept the similarities.

To me, the main factor is that people can opt out of Web3 and can only lose what they put in. But with AI, you can lose your job and your entire life can change regardless of whether you participate — and you can’t opt out! To me, the negatives of AI therefore greatly dominate the negatives of Web3, which is limited to people voluntarily investing in things. The negatives of AI even include a 20% chance of humanity’s extinction according to most of the very AI experts who built it and the AIs themselves.

And yet global competition makes it nearly impossible to achieve the kind of coordination that was sometimes achieved in eg banning chemical weapons, or CFCs globally.

Given this, why are so many rational people on HN much more bullish on AI than Web3? Because they consider and compare the upsides only. But the upsides might not matter if any of the downsides come to pass. Everyone having swarms of AI agents means vanishingly small chance that “bad guys” won’t do terrible stuff at scale (that I don’t want to mention here). By contrast, if everyone has a crypto wallet and smart contracts, the danger isn’t even in the same stratosphere.

MSFTs business is not speculating on BTC, which is why they did not speculate on BTC.

Good businesses make bets that turn out to be bad all the time.

And it remains to be seen whether this bet will turn out to be bad or not.

I honestly feel anyone comparing AI to Web3 is absolutely delusional. Web3 hasn't had a single mass use-case except being a speculative asset. AI is already being used in lot of real life applications (radiology, law, self driving, etc.)

The negatives mentioned for AI here can be used for any technology application which reduces manual labor. AI is gonna enhance your job, or going to displace to a better job for you. Why do you wish to continue doing the work that technology can do 10x better?

Precisely. This is like saying that "the business could have just bet $1B on black at a roulette table and doubled their money!"

Gambling and productive investments are not comparable.

I feel that every response to this criticism on HN always has exactly the same form.

1. Say that AI has many more upsides than Web3, ignoring the downsides.

2. When mentioning any downsides, just say a generic cookie-cutter thing of the form "this was already possible with X", whether X is human activity, or previous technology.

Massive job loss? Was already possible with amazon turk and outsourcing jobs. AI is exactly the same, just a "slight" difference in scale. Nothing to worry about.

Bad actors using AI swarms at scale? This was already possible, albeit not with that scale, by -- um -- botnets and maybe crime syndicates. So, once again, don't worry.

My whole point is to look at the downsides and note that the only losses possible in Web3 were of money voluntarily committed to it. While people can opt out of any harms by Web3, they cannot opt out of harms by AI. This is a major deal, and you dismiss tons of warnings by the very experts who made it.

> Massive job loss

I feel that people become too entrenched in the way things are (people need jobs to make money to live) and lose sight of the bigger picture: machines doing the work that humans have to do now should be a good thing. That it would not be, because in the current system it would result in a small number of people having great power and wealth while the majority of people have little, means the system should change, not that we must not develop the technology.

If you think that the development of AI to take people's jobs and concentrate power is coming and is bad, then you should want to change the system, because that is what the system is encouraging. If you think that the development of AI to do people's jobs for them and unburden humanity is coming and good, then you should want to change the system because it is not set up to gracefully massive unemployment due to automated efficiency gains.

If you think this whole AI thing is a bit of a nothingburger and not going to have the broad impacts that are being speculated, well carry on then.

The curiosity of humans and drive to create new things and uncover new knowledge is universal, and stronger than any society or culture has proven to be. Technologies destroy societies that don't adapt to them, societies don't destroy technologies that they don't like.

The current system is capitalism.

People view handouts as "socialism and bad", even UBI.

Under the current system, in order to get money, you have to do work that is so useful to some client, that they will pay you. Half of all Americans are working for corporations. They don't want to be out in the market trying to sell their services. They want stability so they can feed their family. And women want men to have a stable career etc.

The cascading effect of people who get laid off and are told "learn to X, LOL" will overwhelm X, it's like rats from a sinking ship.

Jeffrey Hinton said it the other day -- your utopian vision should help people, but we live in capitalism. So it will do the opposite.

And any historical analogies to what humans did in the past to adapt to challenges and competition are not really applicable because now AI will be far smarter than humans, and at better at organizing. And it will be deployed by governments and corporations which already have most of the power. Individual humans adapting could be as quaint as, say, horses adapting when cars were invented, or oxen adapting when tractors and combines were invented. The adaptation was to breed less horses and oxen. How's the horse population doing today?

I'm not making any analogies to the capabilities of previous technology. I believe it's going to seriously shake things up. I'm saying that the correct response to seeing such power on the horizon is to prepare society to harness it, so that as many benefit from it as possible, rather than be destroyed by it. There is no evidence that people are capable of stopping the technology train, it has never happened.

I am saying that if you worry about the dangers of AI, the rational course of action is to spend your efforts orienting society to have the best chance of benefiting from it, rather than spending your effort trying to prevent its development.

Or course it has happened, otherwise why don’t we have chemicals weapons, nuclear proliferation and CFCs everywhere ?

Now you've opened the "comparison to previous technologies" can, not me.

For chemical and nuclear weapons, it makes more sense that they are restricted, because they are explicitly weapons. They have no benefit. The technology that underpins nuclear weapons does have benefit, and there are many nuclear reactors in the world. I don't know very much about chemical weapons, but I am guessing that the same chemistry discoveries that enabled chemical weapons has also gone into making useful chemicals or medicine.

For CFCs, we realized the negative impacts, and successfully internationally coordinated to stop using them. This happened after it became clear that the real harms that were actively occurring were not worth the benefits.

You're comparing two completely unrelated things and phrasing it as if we have to choose one or the other.

If your aim is to criticize the risk of AI, you'll have plenty of supporters here. Adding web3 to the conversation is unnecessary and you're going to get called out for arguing a false dichotomy.

Reply on news.ycombinator.com