> OpenAI defines [AGI] as "highly autonomous systems that outperform humans at most economically valuable work."
Well in that case: Excel reached AGI decades ago. It outperforms humans at the single most economically valuable task in the world: adding up numbers in a spreadsheet. Congratulations, Microsoft :)
And I would argue LLMs aren't either. A lot of economically valuable tasks involve manual labor. Can LLMs fix your plumbing?
It would help if OpenAI would define what 'most' means and give some examples.
Remember they said their LLMs are like a phd level brains? It's just hype and bullsh!ttery. Don't get me wrong, genAI is great and really useful, but it's not AGI imo. And there is no consensus what AGI means, so the industry should come up with an agreed definition first.
[0] To me autonomous means self directed, e.g. we humans look at things and then decide which problem or thing to work on. LLMs don't do that. We have to tell them. Yes they might work on a sub task autonomously, but still require human input later on (review), because we can't fully trust its output is correct.
I think Demis Hassabis has a better take on AGI and he believes it's still years away.
----
"Demis Hassabis says current AI falls short of AGI. Solving hard problems or producing impressive outputs does not prove a system has broad, reliable human-like intelligence. AGI would need creativity, planning, long-term memory, and consistent cross-domain reasoning—qualities current models lack."
However, I am not sure why European mass surveillance and imperialism are supposed to be any better than those of any other country. Somehow that never gets answered on sites like that.
simply put, they aren’t the big bully. they don’t have the worlds strongest army and control of the most important currency to strongarm pretty much anyone who stands in the way.
None of what you say makes Europe better when it comes to the topic at hand, though?
If I have to choose US or EU to host my website, the size of their army, the strength of their currency, or their history as political bullies doesn't affect my blog or nextcloud instance.
What affects me is their privacy and data harvesting laws. EU is nominally better with the former, but the latter is eroding very fast and they're roughly the same.
> but the latter is eroding very fast and they're roughly the same.
Citation needed. Some politicians keep on pushing for data harvesting and secret backdoors and what not but most attempts have been blocked and we're still nowhere near the level like, e.g., in the US.
The street isn't the problem. Cars and trucks are. If those streets were filled with people walking, cycling, etc., the health penalty of living next to one would vanish overnight.
EVs aren’t a panacea but they’re a lot better than a 10% improvement. You still have tires but losing engine particulates, fumes, noise, and heat is a huge win in a city - bike or walk near one and you can feel the difference.
Generally, the busiest streets are pedestrianised; you just can't move that volume of people in cars.
(Actually, the _very_ busiest streets are probably major bus gates (where, again, there are no cars), those can have fairly ridiculous density at peak times.
Why not? A lot of streets often seem busy simply due to the incredible space inefficiency of cars. What I like to do sometimes, when I wait at a traffic light with my bicycle, is to count the actual number of cars that cross during a phase. The amount feels large, but when you count, it's often not more than the amount of cyclists that cross in the next phase, and they take up just a tiny fraction of what cars need.
Sure it absolutely makes sense. But for me and my eyes 200 people passing by are somehow still less "busy" than 25 loud and stinky cars. That's what I mean with I need better wording
Bingo! Any country borrowing in its own sovereign currency cannot default on debt in that currency, unless it actively decides to do so for political reasons.
This is a fantasy. Do you really think lenders would just not notice if America inflated its currency away to get rid of its debts, and they'd just say "aw shucks you got us. Anyways, here's a new loan at the same terms as last time." ?
And what about the American public? Look at how much Americans freaked out over a year or two of 6% yearly inflation. How do you think Americans would respond to 30% monthly inflation like in Argentina or Turkey?
A government with its own sovereign currency doesn't need lenders. In fact, it is the government that allows lenders to lend money - not the other way around.
> Anyways, here's a new loan at the same terms as last time." ?
Why would they need to get a loan for from a lender?
> How do you think Americans would respond to 30% monthly inflation like in Argentina or Turkey?
Do you think that printing money automatically causes inflation? How does that work? Do people have some kind of magical device, or is it a physical phenomenon that can be measured? Of course not. Printing money only causes inflation if it is spend in a way exceeds the capacity of the market.
$1,000? According to ccusage, I used around 3,200 USD worth of API credits last month, but I'm on a plan that only costs around 100 USD per month, and I'm not even a heavy user. At the end of each week, I have typically used about a third of my weekly Claude limit. So either their APIs are heavily overpriced, which seems rather unlikely, or they subsidise subscriptions to the tune of 100x or even more.
Why do you think the API being overpriced is unlikely? Seems pretty likely to me that sub sells at cost and api is the massive markup they force on enterprises.
But since the randomness is obviously borked, it was much better to use v4 and find out about it after just 15K records instead of X million records later.
> but I don’t think violence is funny or justified
Well, that's okay, because even Sam Altman disagrees with you. He absolutely believes that violence, including deadly violence, is justified - hence his contract with the US Department of War to use their systems in kill chains.
Perhaps the problem is that whoever threw the cocktail didn't use AI to select him as a target, or maybe he didn't receive payment for throwing it? Because what other difference is there?
I mostly agree with you - he seemed happy for the chance to play the victim. When the system is working, war is different because it has democratic process behind approval (Iran is obviously showing the system is breaking down)
But just because horrible people exist in positions of power doesn’t mean I have to become horrible myself. I accept that there is a threshold where that changes, but I think we would disagree that we’ve hit that threshold. If anything violence now just gives more excuse to justify further consolidation of power (look I got attacked! The anti AI people are crazy, any criticism of me is just encouraging them!) Imagine if it was a serious attack on sama, they could spin it into some serious gains for them.
Could you explain how the Vietnamese were involved in the US democratic process that resulted in around 3 million of their people dying? Similarly, how are the Iranians currently involved in the US democratic process to veto the use of AI targeting against them? As a German citizen, how can I object to being surveilled by OpenAI products used by US agencies?
It turns out that those affected by this are actually excluded from the process by design.
One of the more curious perks of being a democracy seems to be that you can also democratically (within your own country) decide about the fate of people in other, nondemocratic countries and then get to enforce those decisions by military...
I don't think that OpenAI necessarily enforces or fundamentally respects the democratic process. After the recent Pentagon spat with Anthropic, OpenAI did not change their stance to conditionally demand lawful usage of their product.
OpenAI can market democratic values very easily, I'm sure the White House loves that kind of dog-and-pony show. But it's pretty clear that OpenAI does not genuinely care about Rule of Law, let alone preventing humanitarian disasters from citing ChatGPT as their abettor.
