Long before this Matt video, I've done just that. A generator of devilish no-image jigsaw puzzles :-) which I indeed wanted to do a little Show HN someday. I made a small python script call a SAT solver to create a MxN puzzle with only 1 solution but only a few jig shapes, so maaany 'almost solutions'. Then I laser cut a plastic board and made it real, and gave one copy to a friend. At least a nice use of SAT solvers that is more efficient than Matts brute force approach.
I have a cute somewhat related story. I wanted to make really unique rings, so I decided to make them from some random metal from the forgotten realms of the periodic table.
I asked for a site that sells many rings of many different metals, with no luck. THIS metal, it turns out, is really difficult to cast. After many months with the idea shelved, one day my girlfriend found some chinese company that managed to do them. US$ 2k rings, and ugly as hell... but at least unique!
The girlfriend soon left me, but, well at least I still had the rings... until a few weeks ago when I noticed that I lost them too. C'est la vie :-)
When researching for this project, I did look up the densest metal possible: Osmium. Osmium has density of 22.59g/cm^3, which is 17% greater than gold's 19.28g/cm^3. (For comparison, lead is 11.34g/cm^3 and silver is 10.5g/cm^3.) Sadly Osmium melts at 5500ºF and can form toxic vapors, so making a ring out of it would be impractical.
Iridium has almost the same density as osmium (indistinguishable without precise instruments) and unlike osmium or rhenium it is neither toxic nor radioactive (even platinum is very slightly radioactive).
So iridium would be the safest among the densest metals for contact with human skin.
Unfortunately, not only iridium is very expensive, but it is extremely difficult to shape, being much harder and much less ductile than platinum (and its melting point is also much higher).
Platinum-iridium alloys are much easier to shape than pure iridium, but still much more difficult than pure platinum. In the second half of the 19th century, when new standard meters and kilograms were made for distribution in all countries and to replace the original standards made of pure platinum, an alloy of platinum with 10% iridium was chosen, as the most resistant metal to mechanical and chemical degradation that was known at that time but for which it was still possible to process it into a given size and shape.
As a sibling comment put, the Scott Aaronson post has lots of interesting questions about this. Do the aliens who are watching us being simulated inside matrix think we are actually conscious? What if they freeze the program for 100 years, or run the computation encrypted, or if the 'computer' is just a human inside a room shuffling papers?
Is the computer simulation of a water drop wet ?
I found this article [0] very insightful, where they basically propose that consciousness is relative to whom you ask. We inside the simulation may attribute consciousness to each other. The aliens running it may not. What is relevant is the degree of isomorphism between our simulated brain processes and their real ones. So things will advance from all these back-and-forth nebulous arguments only when neuroscience becomes able to explain mechanistically why people claim to be conscious.
I've been working on computatioinal modelling consciousness and came to similar conclusions: there is a continuum between patterns of matter that have consciousness (humans and other animals, maybe a biological enough computer, etc) which leads to all sorts of crazy stuff being possible. Evolved human ethics and feelings of care are incompatible with these amorphous extended possibilities. It can lead to some ultimate copernican revlolution that ends human exceptionalism, to outlaw consciousness tinkering (for how long?), or to put our heads into sand.
I have entertained the idea (if you believe in the quantum many worlds) that everyone will experience themselves as aging to infinitely old age, because in the parallel worlds, well, you don't exist.
If you want a grim take on this looking into the quantum immortality thought experiment. I can't remember who wrote it, I think maybe Chalmers, but the idea is that while the experiencer is immortal, they will still accumulate injuries, old age, etc. It would not be a very pleasant immortality.
Saying that something will happen across the multiverse is stating a certainty, so long as it's physically and logically possible. Every absurdly unlikely thing will happen in that sense. Everybody will, somewhere (and somewhen) become purple with flashing yellow spots. That's very different from saying "I expect to experience this happening", which is a statement about what's probable.
There was this numerical calculus class at Uni where the teacher forbid us to use the calculator. So I just programmed the integral on it, got the partial steps, and just wrote random numbers to fill the the substeps. Got full grade :D
The other case everybody got to pass the class, but after vacation we found the stack of exams completely untouched under a desk. The teacher had a side business to run...
I've thought a lot about exactly these limits to AI. The reliability problem will be solved sooner or later. Same for tacit skills that somebody trains into AI (manage people, do laundry). But if your job is specifically about being human ("nostalgic jobs", judge, athlete, surrogate mother...), or some niche tightly guarded ability (how to breed some special spice, how many oil barrels ship thru Suez Channel), you are safe - until some unforseeable rearrangement of the economy, at least.
> until some unforseeable rearrangement of the economy
I mean, telling the majority of the population that hey, you don't have secret knowledge, now you are mostly worthless is a completely foreseeable rearrangement of the economy, at least in the sense the rearrangement is coming. The question of how the new arrangement will look is unpredictable and likely spans anywhere from UBI to guillotines.
That's funny, because I'm a PAID subscriber to The Economist, and on opening the page, there is a big ad banner on the top and other on the right, filling more than half the space. (I don't bother using adblockers,though). Newspapers have always shown ads even for paying consumers. As the very article ends with: "Users who pay to block ads in some areas are still likely to find them popping up in new ones."
As a non paying ublock origin user, I don't see any ads. The ad-free internet has already been here for a long time if you are willing to spend 10 minutes install ublock origin. Paying is actually worse, because it shows the fact that you have some amount of disposable income and are willing to pay for things, which greatly increases the value of advertising to you specifically.
I have never really used an ad blocker for any kind of time. I've tried on multiple occasions but then just range uninstall it whenever some random site breaks because of it and I don't realise it's because of it. Last time it was some online shop...
Regardless of the morality / philosophical questions, the modern web feels to janky and unstable for me to trust the house of cards staying standing if you throw some Adblock in there.
(But maybe I just didn't tweak the blocker enough...)
If a site breaks from adblocking it wasn’t worth my time to begin with. If an online shopping site breaks I look at it with a heavy side eye and am thankful I didn’t provide my credit card information to such a poorly engineered and secured service.
I'm a subscriber with UBlock and I still get a "Give a gift" footer. And when they deliver the print magazine, there's ads (including on the back cover) AND a flyer "give a gift." Which is annoying because the flyer always drops out.
Subscribing merely turns the "Please subscribe yourself" ads into "Give the gift of knowledge to a friend."
For real, that comment made me sad. Whenever I see the web in someone else's computer without adblockers I wonder how they didn't get off the internet altogether.
I was thinking on buying/building one robot like this to experiment. Can one fit more than one manipulator on it (so it could more easily do things like hang out clothes, or hold a glass of water and add ice)? More generally, how does the AI model would adapt to diverse robot morphologies?
I can think of two ways:
1. The model has to be built accounting for the extra appendages and input. This isn't very extensible. Each new configuration would have to start from scratch.
2. There are two types of models - first one for controlling the part and the second one orchestrating all of the first types. They would have to 'talk' to each other to be coordinated.
I have for some time planning to do a 'Wikipedia for AI' (even bought a domain), where people could contribute all sorts of these skills ( not only 3d video, but also manual skills, or anything). Given the current climate of 'AI will save/doom us', and that users would in some sense be training their own replacements, I don't know how much love such site would have, though.
