planetsprite

99 times out of 100 whenever I see a new tech demo that absolutely blows my mind and makes me optimistic about the future of technology it's always some big model deep learning AI thing. When are we all going to admit that ML/AI is the final and ultimate paradigm shift of our time

I get the feeling it was outsourced to a content farm, which outsourced itself to GPT-3

Is this costing you OpenAI tokens for every new word you generate emojis for?

The solution to failures of AI in heathcare is transparency of data. OpenAI's models work because they have virtually unlimited data to train on. The scale of training data for doctor bots is one millionth the size. Different countries, organizations, universities need to be as open as possible sharing and collaborating, realizing improvements in medicine benefits all of humanity with almost no downsides.

I think that meme strikes a chord in people because there's an inherent relatability to being stuck in a hellish process almost deliberately designed to be as long and soul-crushing as possible. Some websites/user interfaces, "ecosystems", feel like they are Mr. Bone's Wild Ride in cloud-compute platform form.

If "platform" companies like Apple, Google, Microsoft weren't capable of advantaging themselves via anti-competitive practices, their stock values wouldn't be nearly as high. Their implicit leverage over the markets they oversee via their OSes is the core of why they are the biggest companies in the world. Fines are a cost of doing business, hardly a deterrent when your product (your OS) is the conduit through which entire industries operate.

Every chess player plays a lot of games, Hans' have just been more public/documented ones. You really don't think that every 2400+ player is treating chess like a full time job, and Hans just simply is just outworking them?

His rate of improvement, assuming no cheating since his ban in 2020, is a faster ascendance than any other player in history, including Bobby Fischer. So yes, assuming he suddenly quit cheating cold turkey, his performance would indicate he is a once in a generation talent and could easily become the best in the world, according to the Chess.com report.

predictable joke

People's role in society shouldn't be determined by their biology, but the manner in which their biology manifests in their life outcomes is determined by their biology.

>probably fiction it's literally a fictional story isn't it? Why the "probably"?

Any law where the penalty is a fine only exists for the poor. Any regulation where the penalty is in the millions only exists for small businesses.

As goes the value-based conception of deservedness in America. Your worth == your net worth.

Why should the capital owners, who are providing money and benefits to their employees cede further their assurance those employees are infinitely replicable? They are, aren't they, without the artifice of a union.

In China there's the notion of the Mandate of Heaven, a philosophy that whoever is currently ruling is the ruler justified by god, evidenced by the fact that the universe, or god, has allowed that person to become the ruler.

The general American capitalist philosophy applies this on an individual level. One's impoverishment or lack thereof is a justified realization of their inherent worth. If they were worth more, they wouldn't be employees, they'd be founders, calling the shots, right?

Even if AI got to the point of perfectly passing every expert-level Turing test your degree of rigor as to what "thinking" is would never truly permit any belief of AI having struck the golden nugget of intelligence.

Imagine if we were all self-replicating computers, and certain members of this silicon race began experimenting with making creatures with carbon macro-molecules to create organic intelligence, you could make the same claim in the other direction:

"There has been no step-change advancement in Organic Intelligence in, perhaps, 50 years. All we see today is a product of cell count, in neurotransmitter chemistry able to compress TBs of experiences into c. 300B neurons."

Language models aren't built for math. Their improvement/training cycles aren't sensitive to the exactness and rule-based nature of mathematical language, plus there are probably a lot of bad/misleading examples of math in the source data.

You'd have to be unrealistically pessimistic to call what GPT-3 and other huge language models produce "nonsense".

Amazing a website/company/man whose claim to fame is a service that lets people watch videos and pictures in a virtual rolodex that happened to do a better job at letting people upload videos and pictures in a virtual rolodex assumes it can tackle 100x technical challenges better than any other company. All they have is capital, though I guess that's really all you need.

No human will ever enjoy the feeling of a giant block on their face. Even if the quest was 100x higher res it will never be "default" since it limits peripheral awareness, motion, causes eye strain.

I get Zuck's philosphy on upping the human-computer paradigm to the natural next step but imo horizontal screens at a comfortable 1-10 feet from your eyes will remain unchallenged until seamless brain controlled holograms become a thing.

The same argument could be made for atomic divisions of time à la Achilles and the Tortoise, aka Zeno's paradox. However is time atomic?

Human thinking naturally aligns towards authoritarianism, winner-takes-all, strict unbreakable hierarchies. The only reason democracies are so common in the world nowadays is that they end up being so successful, even the minority accidental democracies end up having a disproportionate influence and growth over the span of history.

81. Video games will get more realistic

I'll amend my statement

It is easier to prove a positive since it requires only the "smoking gun" to be made apparent. It is much harder to disprove a negative because it requires hypothetical smoking gun that could have caused the effect (Han's beating of Magnus) to be disproven. Since we don't have a record of every electromagnetic and sonic wave which passed through the room the day Hans beat Magnus, disproving the cheating claim is likely impossible.

There are two things that must be understood before making any judgement on this case:

1. It is solidly within the realm of possibility that a person who spends all his time devoted to succeeding in tournament chess would be able to devise a cheating method that cannot be easily detected by standard protocols

2. It is impossible to prove a negative, so it is impossible to disprove Magnus' claims

With these things being said, it is both possible that Hans didn't cheat and we'll never know for sure, or that he did cheat and we'll never know how. Given these facts, the most productive analysis would be to find just how unusual Hans' performance, both in the game and prior, really was.

A few things could be determined:

1. How many other high level players have a history of cheating? Have any players who have once cheated in a lower stakes match later been proven to have cheated in a higher stakes one?

2. Based on improvement per game played, how unusually rapid was Hans' development in the past 2 years? Are there other players which have progressed as rapidly as he has despite having progressed slower earlier in their careers?

3. Do all the engine correlation analyses that implicate Hans not fire warning signals when analysis any other game by confirmed non-cheaters? Do they signal cheating in a similar way when analyzing games of proven cheaters?

Is grain actually an aesthetic boon to digital art, or does it seem that way just because we associate it with printed work, and we associate printed work with higher quality.

Here's a more detailed map which shows the actual levels for each part of the city in PM2.5 https://s.hdnux.com/photos/65/37/61/14023807/10/1200x0.jpg

As is the case for 90% of public services in the United States.

The experiment ignores the fact that the heat and pressure applied to the winning M&M's weakens them for future battles.

Yes, this site skews heavily not just in favor of engineers, but the more old-school, "give me the box of tools and leave me alone" style engineer. 99% of humans who use a computer just want a single button that implicitly means "give me content" and they're satisfied.

I'm comparing the number of preventable deaths in the US vs countries with similar GDP/capita

https://www.oecd-ilibrary.org/sites/3b4fdbf2-en/index.html?i...

The US has a mortality rate due to preventable causes comparable to Poland/Slovakia despite having nearly a 4x higher GDP per capita. Even poorer countries with better systems do far better w.r.t preventable mortality. This amounts to an extra 100,000 deaths per year approx vs countries like Italy, Germany, Switzerland, Sweden, etc.

I wonder what percentage of work in the US healthcare system is completely unnecessary from a general perspective but made necessary deliberately to justify the unethical system that allows millions to die unnecessarily.

I think it's the fact that you can subtly pick up the way Dall-e has begun to understand the similarity between different abstract forms. Dall-e has likely intuited that there is a general similarity between the faces of different animals, or a general similarity between different organic objects and different artificial ones. We are seeing the expression of those connections the model has made: compression in abstract representation that triggers the same uneasy feeling human brains get when on a bad trip and they see faces in clouds and shadows.

The training of these image generation models minimizes the amount of necessary new visual information it needs to remember with each new image it sees during training, so whenever it sees a new image-caption pair it abstracts its qualities into an amalgamation of what its already seen. This is how Stable Diffusion managed to fit the complete corpus of human visual memory into 4.2 gigabytes.

Everything is more closely connection than the picture would imply at first glace, like how in Super Mario Bros. the bushes are just clouds colored green. Compression and efficiency leads to a hyperconnectivity in the way DALL-E ties prompts and images. This is why those old deep-dream images where everything looked like a dog's face were so unsettling. This triggers an uncanny response in the human brain, because it sees the vague shape of a face in a tree, or the vague shape of a body in the structure of a building.