Hacker News Reader

Comment by chx | Hacker News Reader

> When the next wave of new deep learning innovations sweeps the world,

that won't happen, the next scam will be different

it was crypto until FTX collapsed then the usual suspects led by a16z leaned on OpenAI to rush whatever they had on market hence the odd naming of ChatGPT 3.5.

When the hype is finally realized to be just mass printing bullshit -- relevant bullshit, yes, which sometimes can be useful but not billions of dollars of useful -- there will be something else.

Same old, same old. The only difference is there is no new catchy tunes. Yet? https://youtu.be/I6IQ_FOCE6I https://locusmag.com/2023/12/commentary-cory-doctorow-what-k...

View on HN

Crypto currencies has the potential to grow the world economy by about 1-3%, as banking fees go down. Add other uses of crypto may double or triple that, but that's really speculative.

AI, on the other hand, has a near infinite potential. It's conceivable that it will grow the global economy by 2% OR MORE per MONTH for decades or more.

AI is going to be much more impactful than the internet. Probably more than internal combustion, the steam engine and electricity combined.

The question is about the timescale. It could take 2 years before it really starts to generate profits, or it could take 10 or even more.

>Crypto currencies has the potential to grow the world economy by about 1-3%, as banking fees go down.

Bank fees don't disappear into the ether when they're collected, so I doubt they have this much affect.

Oh, made my very first retail purchase with Bitcoin the other day. While the process was pretty slick and easy, the network charged $15.00 in fees. Long way to go until "free".

> Bank fees don't disappear into the ether when they're collected, so I doubt they have this much affect.

1-3% was intended as a ceiling for what cryptocurrency could bring to the economy, after adjusting for the reduction in inflation once those costs are gone.

He's saying the fees aren't burned like with mining, so they don't hurt the economy by the amount of the fee: the profit portion of them goes into other investments. The fees hinder parts of the economy making some transactions nonviable, but they don't fully translate to "friction" making waste heat so much as something more adiabatic that goes back in. It's largely an extraneous spring in the system, not a damper.

That's why I stated that 1-3% was the ceiling

I agree with you on the AI point, but with crypto not all is what it seems.

Yes you may have short term growth, this is solely due to there being less regulation.

Despite what many people think regulation is a good thing, put in place to avoid the excesses that lead to lost livelihoods. It stops whales from exploiting the poor, provides tools for central banks to try avoid depressions.

Costs wise, banks acting as trust authorities actually can theoretically be cheaper too.

Well, I agree with all that. The 1-3% was meant to come off as a tiny, one-time gain, and an optimistic estimate of that. Not at all worth the hype.

Basically, crypto is more like gold rush than a tech breakthrough. And gold rushes rarely lead to much more than increased inflation.

Absolutely

Do you have a source for any of these numbers or is this just your speculation? I haven’t seen any estimates from well-known institutions that reference any of the numbers your are pointing to.

All crypto"currencies" with a transaction fee are negative sum games and as such , they are a scam. It's been nine years since the Washington Post admittedly somewhat clumsily but still drawn attention to this and people still insist it's something other than a scam. Despite heady articles about how it's going to solve world hunger, it's just a scam.

This round of AI is only capable of producing bullshit. Relevant bullshit but bullshit. This can be useful https://hachyderm.io/@inthehands/112006855076082650 but it doesn't mean it's more impactful than the Internet.

I agree, 1-3% was a best case. While I agree it's a net zero, even those who argue for it really don't claim much more than a couple of %.

I actually expected objections on the opposite direction. But then, this is not twitter/X.

The point is that something that can easily generate 20%-100% growth per year (AGI/ASI) is so much more important that the best case prediction for crypto's effect on the economy are not even noticeable.

That's why comparing the crypto bubble to AI is so meaningless. Crypto was NEVER* going to be something hugely important, while AI is potentially almost limitless.

*If crypto had anything to offer at all, it would be ways to avoid fees, taxes and the ability to trace transactions.

The thing is, if crypto at any point seriously threatens to replace traditional currencies as stores of value in the US or EU, it will be banned instantly. Simply because it would make it impossible for governments to run budget deficits, prevent tax evasion and sever other things that governments care about.

LLM is not AGI and there's no way to AGI from LLM. Put down the kool-aid.

I never claimed llm's are agi. Not all neural nets are llm's.

AI? Yes.

LLMs pretending to be AI? No.

What you call "AI" is generally named AGI. LLM's are alredy a kind of AI, not just generic enough to fully replace all humans.

We don't know if full AGI can be built using just current technology (like transformers) given enough scale, or if 1 or more fundamental breakthroughs are needed beyond just the scale.

My hypothesis has always been that AGI will arrive roughly when the compute power and model size matches the human brain. That means models of about 100 trillion params, which is not that far away now.

> We don't know if full AGI can be built using just current technology (like transformers) given enough scale,

We absolutely do and the answer is such a resounding no it's not even funny.

Actually, we really don't. When GPT-3.5 was released, it was a massive surprise to many, exactly because they didn't believe simply scaling up transformers wouldn't end up with something like that.

Now using transformers doesn't mean they have to be assembled like LLM's. There are other ways to stich them together to solve a lot of other problems.

We may very well have the basic types of lego pieces needed to build AGI. We won't know until we try to build all the brain's capacities into a model of size of a few 100 trillion parameters.

And if we actually lack some types of pieces, they may even be available by then.

Reply on news.ycombinator.com