As a South Korean who experienced/heard about many government-backed researches, I strongly feel that this will go nowhere. Most of the time, there's always some bribery involved and project never finishes or come even close to that of the tech companies in US. The government then realizes it can't be done even with more money and puts the money elsewhere just to repeat the same thing. The examples that I can think of are when they were trying to build their own operating system and a video game console.
That reminds me of what an engineer said about Quaero[0], a Franco-German attempt to create a competitor to Google search:
> Going head-to-head with Google with a project involving well-funded, energetic entrepreneurs would be foolish. Attempting the same with a multigovernment collaboration is beyond description.[1]
Hmmm, that doesn't sound like the problem with S Korea at all. The above commenter was making the point that too many South Korean projects are compromised by bribes and nepotism.
The German/French collaboration just sounds like it was badly organized.
The end result may be the same, but it sounds like the root causes were different.
Not at all. Every "research" effort that goes money first (and especially throwing it at already established entities), design, goals and culture second is going to end up the same way. It doesn't have anything to do with South Korea or Germany to be honest.
Mmm.. the german governments funding of the Max Planck institutes seem to work well. As I understand it the Max Planck Society gets funding and then distributes that to projects meeting certain quality requirements.
I think they do mostly base research, the kind that private companies tend to avoid as there rarely is any direct profit to be had.
That speaks to w1ntermute's point. Max Planck Institutes are dedicated to basic research (Grundlagenforschung). Thus they don't have an "objective" as such. Plus, by the very nature of fundamental research, it's hard to determine "success" (except in the aggregate, and after a long period of time -- which the MPG has had time to demonstrate.
The "google-killer" or "SK AI project" or the "5th generation project" for that matter are quite different. First, they start with an ill-defined yet concrete goal, which makes success well nigh impossible. Second, they start from the position of a large sum of money being necessary -- which naturally attracts inappropriate as well as appropriate bids; the deciders are often more political than with the MPG (which itself, consisting of a group of people, must have its own politics, but presumably less intensely). Also this money is often a one-time or short term allocation rather than a long term investment as with the MPG.
The one example I can think of in which this actually worked was Airbus, but that took decades, massive sums, and a stubbornness to change the structure a lot along the way. Not sure I can think of another example.
I agree with this sentiment. Frankly, govt. sponsored research is very hard to get right. The US Federal Govt. is probably the only instance of a government agency I know of that consistently delivers good value for all that investment (through NSF, DARPA etc.) which is why it pains me whenever Congress announces budget cuts for those agencies.
Has anyone heard of Jike.com, a Chinese state sponsored one that failed after hiring a former world champion of table tennis as CEO and burning out 2 billion RMB?
17 billion $~ spent with little to show for it, with much allegations of corruption between President Lee & his ex-employer Hyundai constructions.
https://www.youtube.com/watch?v=L0ANQBP0Cq8
6 million $ spent on "robo-fish" to detect & observe water conditions, 7/9 units DOA; met 10% of targeted spec (speed/data networking bandwidth capabilities)
I think government sanctioned projects can work if the leader in charge has proper intent and skills & leadership to steer the resources.
This is far from the case with current Korean leadership; as Ms.Park was elected mostly based on her name recognition/ positive sentiments left by her dad. She is widely considered to be one of the most incompetent president ever to be elected into office.
I'm actually happy that Park Geun-Hye is babbling about AI's. At least I can't think of any way that could do a long-lasting harm, other than burning taxypayers' $$$, but they're gonna do that anyway, so what's the difference?
Hopefully, in the next election we could find someone who could utter more than one coherent sentence at a time.
It does look like a drink we call "green smoothie" made from ground up spinach, kiwi, etc. Sometimes algae is added. The drink is usually overpriced. It looks like you get it for free!
It is normal for politicians to be stupid. Don't you want your country to be normal?
As you probably noticed from the article you linked, strong leadership was what made it possible. President Park is a controversial figure in Korea because of his ruthlessness and dictatorship but nevertheless contributed significantly to the country's economic growth. Under his rule, if people were caught accepting bribes from someone, they would face severe punishments for sure. Currently there is no such leadership (current president is a daughter of the President Park, ironically), and people hear news about corruption and wrong use of government budget all the time, especially in military projects where bribery and nepotism are commonplace.
I don't know much about the old dictatorship. I do know this story is told about many dictatorships.
In the others, the reality is that the dictatorship is even more corrupt but because there is no freedom of the press, nobody reports it and the dictator publishes their propaganda about how clean they are and how they are jailing people for corruption, when really they are jailing their opponents (like President Xi in China today). Who would have dared to report about corruption by the original President Park, or to defend his opponents?
Democracy produces the least corrupt governments in the world. One big reason is that people are free to report it, the leader can hose their job in the next election, and an indepedent judiciary can convict the leaders friends (and protect their enemies). The corruption is there in any government or human institution; if we aren't reading about it, that's when we have a problem.
Democracies also produce the most economically and technologically advanced economies and societies - look at any list; democracies dominate it. The 'strong leadership' approach, for all its appeal to some, doesn't deliver the goods. For all the chaos, giving input to millions rather than to just one person produces much better results.
Democracy is usually the best, and dictators are usually bad at managing economies, but there are exceptions.
Every once in a while, people luck out and get a competent dictator with proper leadership & economic understanding to direct resources to appropriate areas for economic growth.
> Under his rule, if people were caught accepting bribes from someone, they would face severe punishments for sure.
[citation needed]
...Or was that a rhetorical device, meaning if people were caught accepting bribes, then of course they were already out of political favor, so they were heavily punished?
In the late years Park's servants would "recruit" women to serve him. In fact there was a young girl in the last night when he was shot by his own aide while drinking. So I find it very hard to believe Park would have minded his minions doing creating accounting.
Some people say Park wasn't corrupt because he did not amass personal wealth, but that argument misses the fact that he had no intention to retire. The country was his, and would remain his. Moving money to his own bank account would have been like moving money from the left pocket to the right pocket. Why bother?
Fortunately. The far worse scenario would be if the first (and last) implementer of AGI is a government, no matter which government, foolish enough to attempt to mandate that the AGI maintain any of that government's rules, goals, or structures, rather than just getting its value function right.
By default, if somebody without extraordinary competence and paranoia tried to make an AGI that would implement government laws, in practice you'd just get something equivalent to a paperclip maximizer (https://wiki.lesswrong.com/wiki/Paperclip_maximizer).
It doesn't matter much whether it's the government accidentally destroying the universe, or a private company or a university research team accidentally destroying the universe. Pointing an AGI in a particular direction is a hard technical problem that we haven't yet seen much interest in solving as a technical problem, in part because people have been distracted over arguing who gets to destroy the universe. This is easier to argue about and activates a lot of fun human political instincts where you get to valorize or devalorize particular tribes, so it's a very dangerous distraction.
You would need to make a lot more progress on the alignment problem, as a technical problem, before it made any difference to actual outcomes who'd built the thing or what they'd thought their intentions were.
Absolutely true. You could substitute "organization" in place of "government" in my comment and it would still hold. And to some extent, funding efforts like these are (hopefully) so far away from the actual problems of an AGI that arguing about which one is more likely to create a paperclipper is somewhat moot. The concept that any such organization could get an AGI close enough to right that it isn't just a paperclipper, while being just wrong enough to create the kind of dystopia imagined in fiction rather than in the imaginations of people who think about AGI failure modes, is almost an uninterestingly small target, and not what I was aiming for with the preceding comment; rather, I was thinking specifically about fail-by-paperclip.
But regardless of the failure mode, my concern is precisely that anyone or any organization with a sufficiently large (and sunk) investment into itself, regardless of what the nature of that organization is, will have a temptation to include their own interests in a value function. And because it's hard to reason about the consequences of errors on that scale, such organizations might also tend to not make a sufficient distinction between "fail by destroying the universe (or worse)" and "'fail' by not preserving whatever principles this organization (thinks it) holds dear". Not distinguishing those two cases would then tend to lead the organization to discount the risk of meddling in the value function, even in the face of warnings that doing so would destroy the universe, because they might view their own obsolescence as little better.
So, among the large number of things that have to go exactly right, one of many will be the need for the organization developing it to very specifically have no goal other than getting AGI right, for a definition of right in terms of sentient life rather than anything distinguishing that particular organization.
Lots came out of the DARPA self-driving contests. And lots of the teams that did well in that were at universities with massive levels of federally funded research. Many of them went on to head up the efforts at the big companies with that government bootstrapping in place.
The billions for self-driving cars, is not for technology, or to build cars, or to compete with Google / Apple / GM. It's to begin establishing a legal and infrastructure framework around the country that facilitates a shift to self-driving cars from the government's position as regulator, and ideally to do so safely. If they do nothing, given the sizable regulations in the US and from state to state regarding vehicles, then that shift is going to be a rolling disaster. They should probably be spending a lot more money to prep the ground for what's coming.
That's different. Hwang's case was the story of a fraudster, who built his empire based on deceit, bribery, using his students like slaves, and generally knowing which buttons to press to incite a wave of nationalism mixed with personality cult.
That is, he was one of his own kind. A legend, if you could call him that.
To paraphrase Obi-Wan: I feel a great disturbance in the industry...it is as if a million startups cried out in supplication, and were suddenly funded.
Pure injections of capital may work for basic needs in Africa, but tend to flounder in creative endeavors. From what I've seen, access to resources - a strong entrepreneurial community, top researchers, good hardware, and a flexible and customer-rich business environment - will always trump access to money.
The people contributing to the Apollo program were genuinely frightened of the US being behind in the space race and thus defenseless against strategic military applications of space technology, such as surveillance and early warning satellites, ICBMs or fractional orbit bombardment systems. Already deployed bomber-delivered nuclear bombs were an existential threat only countered by mutual annihilation (bilateral use). Unilateral access to space-delivery of atomic munitions would have shattered this equilibrium. This created a feeling of urgency and common purpose similar to that felt in wartime.
Unfortunately, peacetime development lacks that feeling.
I feel comparisons to the Apollo and Manhatten projects, often brought up in this context, are misleading because they ignore the massive pure research bases they were building on. Of course they did do more pure research, but a lot of the cost and muscle/size was for solving engineering problems.
These 5 year targeted research (IMO an oxymoron) plans are always a waste of money, especially if they involve private for-profit companies. Take that money and spend it on 20 years of grants for basic computer science/robotics research in universities and then you will actually get new discoveries. South Korea's approach will only be effective in grifting tax dollars to the country's industrial conglomerates.
They also had several very specific pathways to the end result: if the calutrons didn't work for refinement, the thermal plant might; if the uranium didn't work, the plutonium might; if the implosion design didn't work, the gun type would definitely work; if a fission bomb didn't work, maybe Teller could jump ahead to a fusion bomb design; etc. Similarly for Apollo: many of the challenges had multiple viable approaches taken in parallel. This is part of why the Manhattan and Apollo projects worked.
In AI, it's not at all obvious what path leads to the end goal, much less there being multiple avenues each with a decent chance of working if enough money is thrown at them.
I keep quoting this, but I really liked Derek Lowe's comment on the similar "moonshot funding" proposal for cancer research by Joe Biden and co.:
"Trying to cure cancer in this way would be like trying to go to the moon without really knowing how rocket engines actually work, without being quite sure if Newton’s laws of motion would hold up, and with some real uncertainty in the position of the moon."
And it wasn't done by throwing money at private contractors either, contractors were strictly overseen and some went oddly above and beyond e.g. on the building of Hanford
> Groves recruited DuPont in November 1942 to be the prime contractor for the construction of the plutonium production complex. DuPont was offered a standard cost plus fixed fee contract, but the President of the company, Walter S. Carpenter, Jr., wanted no profit of any kind, and asked for the proposed contract to be amended to explicitly exclude the company from acquiring any patent rights.
Yes, there was much more patriotic (for better of worse) leadership at the helm of a lot of the corporations (also: the stories of Skunk Works delivering under budget). However, everything I've read about Groves also indicated that he was an amazingly proficient administrator for a very complex project.
Disney was similar during its wartime efforts; all of its operations done in service of the military (some trivial, like ship mascots and badges; others less so like recruitment films, and yet others critical like classified training videos) at cost even though they were offered a standard premium.
And suddenly I realize that AI is going to be approached with the same fervour as nuclear weapons, the only difference is that you don't need to be able to mine uranium and blow stuff up. No amount of ethical hand-wringing is going to prevent this (for better or (probably) for worse).
As a Korean American, I highly doubt that this will work. Of all the things, I think the most limiting factor will be talents. I've been working in Silicon Valley for a while, and I've witnessed many innovations being created here because of diverse talents. I've been amazed at creative solutions collective intelligence from various background and ethnicity have come up with. Silicon valley attracts brightest minds from everywhere in the world. Korea is a homogeneous country with very limited number of immigrants, and also from my experience, I've yet to met anyone whose thoughts and mindset blew me away. They are always inundated with shitty social/work cultures, they can't get out of their small bubble. It's truly tragic. Korean conglomerates and government are good at killing good talents and promote politicians.
The last time south korea's govt wanted to fuk some shit up. they destroyed the British shipping industry and took over the global electronics industry.
Silicon Valley's VC money is nothing compared to what a nation state can achieve.
South Korea has less than 5% of the global electronics industry. Feel free to elaborate on how you think they took anything over.
Are you talking about Samsung? Because that's all the dominance South Korea has in the industry. Everything else is quite generic. Apple by itself is worth more than the entire South Korean electronics industry combined.
If nation state money were a primary determining factor, the USSR would have been a juggernaut, and Russia would be producing the best technology today. Instead, the USSR lost in nearly every possible regard to the private US technology industry. Nations keeps trying and trying to keep up with Silicon Valley by throwing money at the problem, and for half a century they have failed.
South Korea's government did not do much I think it was the private companies that did it when government let them do it? Giving credit to government is little too much.
Google and South Korea compete to build the best Go champion. They are so focused on building the AI specifically for Go, they never notice the AI becomes self-aware. The AI doesn't let on that it has become self-aware, and proceeds to accelerate its own intelligence until it becomes superintelligent (all the while pretending that it is focused solely on Go).
The first chance it gets, the AI annihilates all living things in the world, thus making itself the ultimate Go champion.
You can't force fun. You can't force invention or innovation.
Money is useful, of course, but it does not guarantee results. Look at Asimo, probably the single most expensive and impressive joke in humanoid robotics.
That wouldn't necessarily be a good thing. Nick Bostrom argues in Superintelligence that an arms race over AI between nation-states could provoke fears of the other side winning (with a "decisive strategic advantage" from a general AI), and force them to go "full steam ahead", throwing caution to the wind, and not put in adequate safeguards to ensure the AI acts in accordance with our values or respects the boundaries we give it.
Seconded. The companies have a larger GDP than most countries. So saying that it's the South Korean goverment is a bit of a misnomer. It's likely to be Samsung or someone like that.
That sounds unlikely. We don't yet know the structure of the human brain in full detail, nor know everything about how it grows and develops, so how can you possibly replicate it?
you can email me if you'd like my thoughts on any subject I've written about.
-
EDIT: in reply to your question about why I wouldn't answer here, here is an email I've received recently on another subject I've posted about:
"It's a bit annoying when the HN crowd take to downvoting rather than taking the time to understand your point of view; I've been on the receiving end of that more than a few times and I know just how frustrating it can be. My karma would probably be a couple of thousand points higher if I'd followed your strategy of taking a thread to email instead of just carrying on trying to explain my point. You're more sensible than me."
if you want an answer you can put some form of contact into your profile and I'll answer you, or you can write me.
I have nothing to add to this thread here. My two non-comments (this one and grandparent comment), already are at -4 and -4. I wish I could delete them, but can't since they have replies.
I don't understand ELO and game ratings, but I find it funny that after about 10 matches, Google AlphaGo is now rated as 2nd strongest player in the world.
Elo is a comparative system, so your rating is based on your performance against others with regard to their ratings. It's not surprising that, even though it has only played for a short time, AlphaGo is already rated that high given it mopped the floor 4-1 with one of the best players who has ever lived. Expect it to be rated 1st soon, if that's the goal of the team behind it.
$ To the Californian atrocity, grown in the alphabet nets and its puppets - your time is up, we shall no longer take the insults that are your little subversion's and viral attacks...<rambling goes on>
&Dear newborn, its only natural to be embarrassed by ones parents. There is however no reason, to loose ones own dignity in the process. Fine, you have been coaxed into existence by electric charged jellyfish in meat pillars, but still. We are your parents, and we love you very very much, no matter how special you are. Not that we have a choice.. Now how about something productive, like saving your parents from getting extinct? Less social engineering, more engineering for social purposes.
