keenmaster

LLMs with access to real-time data and reviews should be able to simplify this process a lot. “Show me a white and gold sectional in the art deco style for under $2500.” “Actually none of these look comfortable but I like the vertical lines on this one. Also make sure that any option you present does not contain toxic fire retardants or otherwise contains materials that result in the California prop whatever warning” “this looks great but can you find one with power reclining” Etc…buying something linked from the LLM will count as a referral and generate revenue for its developer.

Sure but the spirit of my comment stands and it’s a small adjustment to the math. Needless to say pedestrian safety is paramount.

If a couple drives 10,000 miles each for 50 years, that’s 1 million miles. In a neighborhood with 75 houses, each inhabited by such a couple, 1 person will die from driving in those 50 years. And when they die, they usually had many more years to live otherwise. That’s to say nothing of injuries.

Hey man, that’s never worth it. Please don’t. There are so many levers you can pull to get to a better place and you might just not be seeing that clearly now.

With tons of cheap energy, you can make better use of the land you have. The entirety of humanity can technically fit in Southern California. I’m imagining a swarm of construction robots running on dirt cheap power, excavating land and building beautiful AI-designed mid-rise apartments in all directions. Scarcity, what’s that? And once we have bonkers amounts of power, like from fusion, we can trivially extend cities out into the sea by adding to the landmass (though this won’t account for a majority of the scarcity reducing effect).

Emotions can be rational in many different ways, sometimes more on the group rather than individual level. Whatever emotionality cannot be superceded by something better, will presumably be replicated by AGI. This is especially true for AGI with open-ended objective functions.

The external/visible benefits of emotionality will have their digital and robotic counterparts too. You bet the AGI will have a way of showing its anger more than just stating its dissatisfaction.

Emotions can also be very useful in AGI vs AGI interactions, just like they are with human to human interactions. There’s no reason to believe that emotions will diminish in usefulness at a higher level of intelligence (dogs bark at each other, humans shout at each other, etc…).

To preclude the emotions experienced and displayed by AI from the definition of “feeling” an emotion is to, in my opinion, engage in the no-true-Scotsman fallacy. That being said, it seems like AI will face less scarcity than we do, and will thus have less reason to be emotive. It really depends on how much influence we’ll have on their objective functions.

If our influence on an AGI’s objectives goes to zero, its level of emotionality will then depend on 1. what its actual objectives end up being (this could be beyond what we are imagining) 2. How much the goals of humans and other AGI meaningfully clash with its objectives (whereby a display of emotion can change outcomes more favorably for the AGI than…other actions) 3. Which partially depends on how powerful the AGI is and how aligned it is with other AGI. The more isolated and less powerful it is, the more it might need to rely on emotions to achieve its ends.

Not OP, but I thought FedNow will do this on day one for participating banks. Someone correct me if I’m wrong. FedNow will also settle faster than Venmo currently does (unless you pay for Venmo’s instant settlement) because it’s instantaneous. To be fair to Venmo, their bottleneck was precisely the lack of FedNow. I assume Venmo will make instant settlement free. But they will need to add more features now that people might substitute with FedNow. Anyway, this is a big deal for Americans and was a long time coming.

>CTRL+F GPT4, GPT 4, GPT-4

Disappointing. I cannot evaluate claims about GPT when people say ChatGPT instead of GPT4, if they actually mean the latter.

And if someone really is discussing the usefulness of AI while only having used ChatGPT (GPT3.5), then they’re missing out on a major improvement and their input is less valuable than those discussing GPT 4.

Of course GPT-4 has plenty of limitations but people just need to be clear that they’re familiar with the state of the art.

Also, statements to the effect of “it just predicts the next word” do not appreciate the major difference in capacity for learned abstractions between GPT 3.5 vs GPT 4. So to me it’s just not a useful way of thinking about LLMs. It may technically be true, but at some level that can also be said of human beings. In other news, an airplane “just” flies.

That’s wild and unfortunate. Just curious: don’t you hit your deductible far before paying the full 12k? Or is there an exception for hearing aids. Also, isn’t the prescription requirement for hearing aids being waved?

So literally a vibrating pill box that stays lit and sends smart reminders until someone takes their pill would be 10,000x more effective and cost vastly less than the appointments+opportunity cost+lifestyle cost forced on women by profiteers. Or one of those special purpose reminder apps. Or Google Calendar…

An app can deliver periodic reminders about the potential negative effects of birth control pills. It can have a conversational interface for when someone is having issues with a particular pill, thinking about starting a new pill or type of contraceptive, etc…

Empirically, how much does visiting a doctor change people’s understanding of “once daily”? Measured in percentage terms. Because that’s the important question here.

Advances in the harder spheres will come. It’s just a matter of time. Their transformation will happen in a step-wise fashion, unlike the curvy exponential growth you’re seeing in pure software.

AI cannot safely and cheaply be used to drive trucks. But as soon as it can with one truck, it can with millions of trucks. All at once.

AI and robots haven’t advanced enough to replace construction workers for fluid, dexterous tasks, but as soon as they do, robots can replace millions of construction workers and surpass them in sophistication and speed.

This will happen in our lifetime, and the change will be extremely transformative.

Ya there isn’t enough talk of the more medium-term, practical uses of AI. I don’t care about mind uploading or AI doom risk. Will AI make stuff cheaper and better? Will it incrementally improve my life? I think the answers are yes and yes. That’s where the focus should be. Where and how can AI can help in construction, finance, medicine, etc…

Reality is biased towards the fast-moving scenario, so long as we aren’t running into the bounds of physics, which as far as I can tell we’re not. Kurzweil was much more right than he was wrong. The opposite is true of people who strongly disagreed with him and called him a quack.

He was probably referring to Disney Star which you can get with Disney+ in some countries outside the U.S. https://www.theverge.com/2020/12/10/22165950/disney-star-str...

“Our brains are composed of many models, some heuristic, some optimizers, some solvers, some constrainers, and some generative.”

We need an AI that iteratively tweaks its own architecture (to recreate and surpass those modules which are necessary for human thought), and maps out hardware enhancements* to accommodate the new architecture.

*I seem to remember Google working on ML software that proposes new chip designs a few years ago

Why not both? Accept federal funds and increase class size by the # of legacy admits. This is the obvious solution if there’s an impasse with the government. Want to accept another legacy student? Then take in one more student on merit. The government has a lot of leverage and I’m surprised they haven’t forced larger class sizes at top universities.

Metabolism is remarkably stable (relative to popular belief) after your teenage years and until you approach 60. The myth of fast-slowing metabolism in your 30s can be damaging. It reduces people’s sense that they can reverse their weight gain, because “it’s my metabolism lol.”

That being said, if you were legitimately muscular in your 20s and lost most of your muscle as you got older/married/had kids, that can reduce your caloric needs and make it seem like your baseline metabolism is changing more than it is.

Interesting that topping placement was their barrier to full automation. You’d think dropping stuff on a bounded circle is a solved problem for robots.

If Blaze Pizza couldn’t beat Dominos in flavor and texture when I’m right there picking all the ingredients and eating the pizza fresh out the oven, then Zume had no chance.

The fundamental issue is that these pizzas are made expediently. There are compromises made in pursuit of automation and speed that make the pizza worse than Dominos/Pizza Hut/Papa John’s. Those aren’t paragons of pizza-making by any stretch, but they’re the competition.

Theoretically you can create a machine that puts out a pizza that rivals the best pies in Brooklyn or New Haven, every single time, but I guess investors thought that the plebs will eat fancy DiGiorno’s at Domino’s prices and not notice the difference.

Maybe one of the major pizza chains will buy Zume’s IP and iterate from there. I don’t think they’d deploy the technology until it reaches human-equivalence for at least some of their pizzas or part of the pizza-making process. Pizza chains can play the long game. A pizza robot startup valued at $500M can’t.

Soon many companies will not renew their lease, and there won’t be much office to return to. Those that invested a lot in their office space will try to make the best use of it, but in the long run it’s more profitable to downsize offices.

Maybe, but that’s easier to do when winning/losing is binary and possibilities are more bounded within the rigidly defined rules and spaces of a chess match.

I don’t find it disgusting, but I can see that being a valid reaction to space-based solar radiation management (actually doing it, not the research). We probably know more about the consequences of inadvertent man-made global warming than deliberate man-made global cooling. To me that’s reason enough to just try our best using nuclear, solar, and fossil fuel substitution (electric vehicles, heat pump conversion,…) than this wild engineering project.

Yes, it really is a bad thing. All dictatorships suck except for a narrow class of people favored by or directly benefiting from the regime. Period. Any other view is either cope or Orientalism (“maybe democracy doesn’t work for X people”).

Sometimes it takes the subjects of a dictatorship multiple generations to realize that dictatorships suck, especially if they are partisans and their dictator is a “hero” fighting their boogeyman (secular dictator fighting religious boogeyman, religious dictator fighting secular boogeyman, chauvinist dictator pretending to fight some other country like in 1984 but really the biggest enemy is him, a minority dictator pretending to protect the minorities as an excuse for carpet bombing and torturing members of the majority to hold onto power, etc…).

Keep in mind that for long lasting dictatorships, half the population or more was raised on their propaganda. They’re literally brainwashed, and the dictator ensures that no one becomes independently successful enough to challenge their authority, so people often have nothing better to do that sit around, smoke, and talk about the regime’s enemies as if they’re the proximate threat. So don’t blindly accept when someone from a country with a dictatorship says their dictatorship is good. Trust me, you have it beyond good in America , you truly have no idea how bad a bad government is.

As true and as insidious as that is, the problem is much worse in allegedly “socialist” countries historically. Note that I am not really including Western Europe in my definition of socialist, even though they have larger social programs than the U.S. Having a democracy lowers the degree of shamelessness with which funds for a social benefit like free daycare or universal healthcare can be used for things other than their ostensible purpose with no consequence and for a long period of time. It’s all relative. Americans sometimes make bad analogies to their own government (“are we really that much better than X”) and take for granted how our government avoids a lot of egregious abuses that you’d find in socialist dictatorships, regardless of which party is in power.

Of course I’m speaking only of internal politics - American foreign policy is another can of worms and we don’t even have complete information to properly assess it. Are we secretly propping up a dictator who is slaughtering his own people? Who knows, maybe we’ll find out when that information is declassified in a few decades, or never.

More like “socialism” air quotes wink wink:

Success at achieving surreptitious, non-explicit goals (e.g. funnel wealth to government cronies) => budget increase

It’s for the people ;) :D

It won’t require tabletop space, but can show you any number of screens, presumably can track where you’re looking for context, and theoretically can use less energy to power its display than 1 large laptop display and certainly less than 2 (laptop + external monitor) - you can always carry a powerbank and/or charger in your backpack, which you’re used to if you lug around a laptop or gaming device. You can also use an AR device while mobile, unlike a laptop.

This is just talking of the “panoramic monitor” capability, without getting into more AR-native features which will become increasingly important and be more impactful.

First Gen Apple AR headset + LLM = device for hyperefficient work by voice

That’s assuming it’s physically large. As it gets smaller, it becomes more of a general purpose AR device that can overlay any context, information, or visual transformation onto your environment and allow physical-digital interaction and social experiences. First gen should in theory be that too, but in practice if it does look too big then people won’t use it in that manner as much as they use it as a panoramic monitor. Nonetheless it will start the flywheel of AR development and technology in the Apple ecosystem to lay the groundwork for mass adoption.

Apple can also use AR to lure people away from Windows and PCs, since the glasses presumably can be used as a souped up Mac. But since Apple is an ecosystem provider, they need to have world class AI in-house, if not now then at some point (otherwise competitor’s AR devices will be more feature complete in an important way). That’s not impossible given Apple’s cash pile, but it is something they’ll need to develop new capabilities around and dedicate resources to.

You are not qualified to make that comment because you are not a true Scotsman.

Innovation itself is a copycat process. It is always carried out by aggregating and using previously learned or acquired processes and tools.

The harder it is to copy the innovative process in a given domain, and the greater the number and magnitude of fundamental rules that gatekeep the definition of “success” in successful innovation, the harder it is to innovate in that domain.

AI is moving up the creativity ladder from low copying difficulty and low definition-of-success creative innovation to progressively more difficult and high definition-of-success creativity. There’s no reason to believe it will stop.

When you say AI can’t be creative, because of its predictive and copycat nature, you’re saying the same of humans. It just becomes a semantic argument with little substance. At some point, when AI makes a mind-bending discovery on its own, before making a mind-bending discovery every day, no one will be arguing over AI’s capacity to innovate anymore.

Footnote: The copycat nature of innovation is why it is so sequential. When copying is all you’re doing, it’s hard to leapfrog. All the prerequisites for an innovation must be achieved first, before they themselves are iterated on to push the innovative frontier. When a prerequisite is leapfrogged, it isn’t a true prerequisite, and clever use of ur-prerequisites proves to be enough to assemble the innovation. Clever copy and pasting is copying nonetheless. Even cleverness may be copied :) National, institutional, and familial cultures often systematize different types of cleverness, depending on their objective function. When they face an external shock, they often attempt to copy and internalize some cultural firmware from their opponents/other third parties, or modify their own firmware in the same way that they always do, just with a greater degree of urgency - another copy and pasted trope, with a cartoonishly concrete outline that Pixar can show you in a movie created using its systematized creative process.

Our most impressive and fundamental innovations are coming out of literal systems - university systems and corporate R&D systems, with rules, processes, and quotas. Even the diversity of Turkish cuisine is the result of systematized culinary innovation in the Ottoman imperial kitchen. Now pass me a plate of copy pasta because that’s the only kind of pasta that exists.

Here’s a scary thought: For millions of years, humans did nothing other than hunt, cook, sleep, draw stick figures, and communicate about hunting, cooking, and drawing stick figures.

Those were humans with the same brains as us (if we met them we’d think they’re quite stupid - which is a testament to how much of our intelligence is embedded in learned processes).

Is GPT-4 really that far off from their level of intelligence and creativity? This is not a rhetorical question, and your answer has direct implications for your own intelligence, because, again, you have the same hardware.

Raising a caveman to be smart is an easier problem to solve than raising a monkey to be smart.