The engine and the warhead are two of the biggest challenges in making a missile, in large part because anything high performance is also going to be spectacularly dangerous to manufacture.
If a human cares about the work, they can often outperform an LLM because they will keep at it until the work meets their standard of quality. Whereas the LLM will guess and then wait to be corrected. As a recent tweet I saw said: it’s amazing how fast the software bottleneck went from writing code, to reviewing code.
I think we’ll need to split the concept of intelligence into the capacity to accomplish a task and the capacity to conceive and prompt a task. If the former is called “intelligence” then LLMs are intelligent.
But what then do we call the latter? I think the idea of an AI that can independently accomplish great things is where people talk about “general” intelligence. But I think we need a label more specific, that covers this idea that successful humans are not just good at doing things, they originate what should be done and are not easy to dissuade.
To be fair, this is a side effect of financial deadlines in general. A business reporting annually will still try to beat that deadline right up to the last second.
In this particular example a single truckload would be less significant annually than quarterly though.
The environmental movement and labor movement are two examples where citizens organize to go up against corporate interests and win pretty regularly and durably.
Most of those folks would not call it lobbying because of the negative associations of the word. “We have activists, our opponents have lobbyists.” But it works the same way.
It is specifically protected in the First Amendment: “Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press; or the right of the people peaceably to assemble, and to petition the Government for a redress of grievances.”
You might want to look into the industry funding of environmental organizations and the decline of union membership before you decide with your whole heart.
Alan Kay had a great quote: “The best way to predict the future is to invent it.” A quote that many technologists and inventors have used as inspiration over the years as they create things.
It looks a little less inspiring when applied to prediction markets. Because yeah, one way to win a bet is to bet on something you already know you can make happen. AKA insider trading.
But even worse, another way to win a bet is to cajole / threaten / bribe someone into lying about what really happened. That’s the stakes in this story. And at that point, the social value of prediction markets has gone negative. You’ve no longer got a tool that helps predict reality, you’ve got a new way to reward people for lying about what is real and not real.
Genetic heritability, which is a subset of heritability, does not generally mean what people think it means, though. To the extent IQ can be attributed genetically to parents, it does not imply a "compounding" effect where smart people continuously have smarter kids and dumb people have dumber kids. This can understood by analogy to other personal traits like eyesight, height, weight, hair color, etc. which vary within a range across generations with occasional unpredictable outliers. This is why folks who do test for intelligence, have to test each person individually and not just rely on bloodline tracking.
Yes, but since the heritability is high, the average IQ of the children will be close to the average IQ of the parents, despite the fact that it will tend to regress towards the mean.
That’s exactly how it works in the standard additive model of heritability, and we have lots of empirical evidence that heritability of intelligence matches that model very well.
Copyright protects copying. Scraping content does not violate copyright if the content is not republished. Otherwise Google and all search engines would be illegal.
It doesn’t hurt SerpApi’s defense against Reddit’s lawsuit, because SerpApi does not need to prove they have your permission. They only need to prove that Reddit does not have the legal authority to prevent SerpApi from scraping your content. Which they almost certainly don’t.
The warnings are nice but he could just say what it is. Members of Congress have immunity for what they say on the floor of their chamber in session, classification or no.
Immunity from prosecution, maybe, but not immunity from consequence. I can’t imagine congressional leadership would think of it as a good look—and isn’t the “need to know” based on the congressperson’s role? For example don’t they brief only congresspeople in specific roles on specific matters, like the so-called “Gang of Eight” on intelligence matters? [0]
It feels a little like keeping the filibuster around: maybe technically it’s within their power to change the norm, but once unilaterally spilling secrets becomes The Done Thing, it’s hard to imagine it wouldn’t spin out into a free-for-all.
For all the mud that gets slung around, I think congresspeople really don’t get there without some kind of patriotic instinct, some kind of interest in the United States’ ongoing functioning. And I certainly can’t imagine they’d keep getting access to new secrets after pulling something like that, one way or the other…
This is all true and it kind of defines the scope of the harm he is talking about: bad enough for vague warnings, but apparently not bad enough to risk consequences to seniority etc. by outright revealing it.
Worth noting his full quote is that people will be “stunned that it took so long” for the info to come out. Which is not quite the same thing as being stunned in general.
> congressional leadership would think of it as a good look
Why do they have any power? Wyden was elected by his constituency. The "congressional leadership" can go pound sand. To the extent they have any power here it should immediately be completely neutered and then removed.
They can remove him from all his committees, including the ones that give him access to this stuff to begin with. If they really work at it, they can freeze him out to the point where he can't get anything done on this or any other issue. And they can use him revealing the information as an excuse to avoid blowback from their own constituents. It's not as bad as in the House, but it's pretty bad. Oh, and they can probably deprive him of the floor the second he starts to say anything "interesting".
Yes, there are serious problems with the way Congress is organized, but there's probably a reason that practically every parliamentary body on the planet has similar problems.
> and they can probably deprive him of the floor the second he starts to say anything "interesting".
So, move the show off the floor, never has it been easier to reach the population as an individual. Are the citizens that enraptured by "the floor" as it is? It seems to me, that if you were serious, this would be no problem at all.
> there are serious problems with the way Congress is organized
None of that is dictated by the constitution. You can change the way committees work overnight if you want. Some would argue that this happened in the 1970s and 1990s when party politics fully invaded what used to be assignments of seniority and experience.
> but there's probably a reason
Corruption. It's worth a lot of money to certain people. You can either design that out of the system or reduce the total power of that system relative to the population.
I'm not sure you can do much until you get down to the bedrock problems here.
> So, move the show off the floor, never has it been easier to reach the population as an individual. Are the citizens that enraptured by "the floor" as it is?
Nope, but his immunity from prosecution for disclosing classified information only applies during Congressional debate. Once he takes it off "the floor", they can just arrest him.
To answer your question, Congressional leaders are elected by their colleagues. Their power comes from that and from the rules that Congress writes for itself.
reply