peterlk

Life and chess are not the same. I would argue that this is showing a fault in standardized testing. It’s like asking humans to do square roots in an era of calculators. We will still need people who know how to judge the accuracy of calculated roots, but the job of calculating a square root becomes a calculator’s job. The upending of industries is a plausibility that needs serious discussion. But human life is not a min-maxed zero-sum game like chess is. Things will change, and life will go on.

To address your specific comments:

> What are the implications for society when general thinking, reading, and writing becomes like Chess?

This is a profound and important question. I do think that by “general thinking” you mean “general reasoning”.

> What happens when ALL of our decisions can be assigned an accuracy score?

This requires a system where all human’s decisions are optimized against a unified goal (or small set of goals). I don’t think we’ll agree on those goals any time soon.

Yep. A ton of companies raised a boatload of money in 21/22 to “ride out the storm”, and I think many of them are zombie firms who will collapse when they run out of the excessive runway that they have. I keep wondering whether this will attenuate any recession and actually smooth out the badness or if it just means we’re on for an extended storm.

And it comes with a 3mm headphone jack! Hallelujah!! I’m so, so glad someone is stepping up to fill this niche. My only complaint, and it’s kind if a big one, is that the screen is too damn big. I’m tired of phone manufacturers forcing me to carry a television in my pocket. I want something small that I can hold in one hand without a pop socket. I’ll trade performance, camera quality, basically anything for a phone with a small form factor.

AI continues to advance (there is mich more beneath the surface if the endless ChatGPT nonsense), the fediverse is gaining more traction, biotech continues to advance. There continues to be immense opportunity. But you have to skill up and look past the tech and business models of 15 years ago.

No it isn’t. “Tech” is only slowing down for a very narrow definition of tech. Perhaps we forget that rather recently the world created a vaccine for a novel virus using cutting edge techniques, that genome sequencing costs continue to fall, that genome construction costs continue to fall, that synthetic meats are on the horizon, that infrastructure for electric vehicles is now present in much of the world. Tech innovation continues to flourish, that growth just isn’t focused in computer hardware any more

Something that often gets overlooked in systems like this what happens when there are bugs. Mechanical control systems are often designed to be durable, and this gives rise to simplicity. Why add all this complexity? What happens when a bug in the software of the bike locks me out of functionality I’ve paid for? This approach is so user hostile.

A bike throttle connected to a carburetor is beautifully elegant. Twist throttle -> pull wire -> wire pulls open air intake valve. If anything ever goes wrong with this system, it’s usually a $5 wire that needs to be replaced (fuel mixture problems are further down in the pipeline.

I do love the irony that these decentralized systems often end up with the problems they were trying to solve. A trusted third party is the opposite of a decentralized system, and this service could be provided by any FAANG.

Ok…. So I’m going to try to rephrase what I just learned from this comment. Can someone tell me how far off I am?

The universe is not 100% empty, so the index of refraction is not 1. This could be explained by lots of phenomena like dust particles or stray radiation, etc. Because of this effect, for practical purposes, light that we see as res “travels slower” because it has traveled through more media than blue light.

Side note about MITs marketing... The original paper [0] was published by 5 authors. 3 from Stanford, 1 from MIT, and 1 from Autodesk.

I find MITs aggressive marketing pretty off-putting. Any discovery related to MIT _always_ has MIT in the headline, even if their involvement was secondary. It's annoying and makes me feel like MIT cares more about optics than good research. If I'm not alone in this belief, MIT will erode their brand by doing so. They certainly have for me.

[0] https://cs.stanford.edu/~rcwang/projects/lego_manual/

Wikipedia actually seems like a good replacement for personal "websites" in the social media era. I'd love to see the wikimedia social network

> not structural or risky behavior from investment banks in 2007

I don't see it that way at all. Over the last couple years, we've seen an explosion in private capital investments, NFT and crypto fads (questionable, but I would argue that these were large enough to be considered structural), greatly increasing home prices, supply chain disruptions, the war in ukraine (and all its knock-on effects). Things are a bit topsy-turvy, and I am hesitant to write off all of these things as "not structural".

I thought you all just had one number for everything! 0118 999 881 999 119 7253

https://youtu.be/ab8GtuPdrUQ

I absolutely love foley in music. I haven't updated it in a while, but I have a playlist of music that contains foley that I like: https://open.spotify.com/playlist/41uMIYCFV2XXaFtPWSNo9V?si=...

I find that adding foley to music makes it easier to engage in active listening of the song. For anyone who has ever produced or thought of producing music, adding foley to tracks is _really_ fun. Replace a snare hit with a capacitor pop or a noise build with an airplane taking off. It really adds depth to whatever you're making. It often forces me to step up my production game too because natural sounds come with all kinds of undesirable frequencies and spiky amplitudes that need to be smoothed out in order to cohesively fit into a track.

This is really neat. But I remain skeptical

> In both instances, a catalytic reaction drives the nanoparticles to produce antimicrobials that kill harmful oral bacteria on site.

So, what happens if I swallow this stuff that I'm putting in my mouth?

Has anyone solved the data problem? I have taken many stabs at this over the last 10 years, and the primary problem is that music is _extremely_ subtle. Even something as seemingly simple as chord progressions are very difficult to annotate properly without deep training. And even then, experts can disagree.

The thing that's missing is the data. If we had midi transcriptions of 100k songs (abc notation could be fine too), we could probably get really interesting stuff, but most of what is available is lossy chord transcriptions and classical music (public domain). So if you want to automatically create something that sounds like mozart, you're in luck!

But this isn't really satisfying to me. For generative music, we're still largely stuck with encoding musical rules in code rather than feeding data to a transformer. To me, the former feels much less like AI than the latter. The data is all locked up behind an impossible quagmire of copyright.

But if I were a sheet music publishing company, I would be seriously considering the future of music creation with AI given my broad access to notated music & metadata (is this an original score, or a grade 1 simplification?). But again, music copyright is a pretty complex contraption.

Ah yes, of course. Thanks for the clarification!

> The scientists say their model of inner core movement also explains the variation in the length of day, which has been shown to oscillate persistently for the past several decades.

How do we measure persistent changes to the length of a day on Earth? This seems like a maddeningly slippery problem.

- Geosynchronous sattelites? What if they're slightly off?

- Star readings? Can we measure that precisely enough from the surface to find tiny fractions of change?

- Any surface-based reading also has to deal with the non-spherical nahure of Earth, and its wobble (the motion that gives us seasons)

How do they do it?

A family member started getting pretty deep in the crypto game, and went to one crypto conference. Afterwards, we hung out and he was really deflated. He just said, "none of these people understand what's going on. I asked what would happen when the underlying asset values went down, and they tied themselves in knots trying to unravel their own cleverness".

Crypto is interesting to me in the way that algotrading is interesting. Lots of interesting math doing fun things, but not worth really doing unless you have _massive_ resources. To me, there is 1 huge potential win for crypto - replacing clearing houses and banking infrastructure. Winning that is going to make a small number of investors extraordinarily wealthy.

There is 1 other interesting case for crypto, which is a common API that doesn't care who the winner is. But that's a questionable win.

Everything else in the crypto/defi/web 3.0 world is just a scaled MLM ecosystem.

I have no involvement in either of these companies, but I'll mention that this seems like a beta version of uberduck. Personally, I think uberdurk is awesome, and probably worth a look

I think you're out of touch with the realities of being poor. Are you familiar with the ghetto tax?

Let me tell you the situation of an old friend. He was working his ass off to put himself through college even with scholarships because his family was too poor to pay for anything and could not qualify for a loan. He worked one job taking care of his special needs brother, and another job at a brewery while getting his degree. Over the summer, he worked three jobs to cover the bills for college. After college he got a job, but still needed to take care of his brother along with his mom (who has no money, dad's out of the picture) - extra cash went to help his mom. His brother could not get a job because as soon as a special needs person has more than 2k in assets, they no longer qualify for help and the extra cash needed to shepherd him through the time with no help is not available. His mom doesn't really make money, but how many people are going to not take care of the single mom that raised them? _Everything_ is a struggle - constantly. It is exhausting for him, and I dare say that most humans would not match his grit. Passive income is a desperate hope to bridge the gaps.

There are plenty of dumbass florida-man people, but writing off poor people for lack of effort is not a useful generalization.

There are already so many comments here, but I would pay $150 for rocksmith built for piano

If this phone allowed for 2 SIM cards, I would pay a silly amount of money for it

My friend once ordered a hot tub with 2 day shipping on Amazon Prime, and has often wondered about the economics of that purchase.

This is a rational argument, and I see it a lot, but it does not probe deep enough. GPT-3 use cases "fall apart at the lightest real world case" because our expectations are wrong. For the vast majority of computing history, computers have been used to do things deterministically and accurately. Next-gen AI does not operate in this way. It encroaches on the human part of work where we expect and allow for people to get it wrong.

GPT-3 does shockingly well at classification tasks with basically 0 training/prompting (outside of the base model). And it works for an incredibly broad set of use cases.

But using QA and generation are much harder to judge because we can't say (in general) whether generated text is "correct"

I'm not quite sure I follow your question, but I have to keep rediscovering that min-maxing life is not a great strategy. Trying to squeeze every bit of advantage out of what I read, how i work, where I travel, etc. is not the optimal strategy for me. Furthermore, the most successful people I've met tend to focus on doing things that they enjoy rather than on min-maxing.

For me, this has meant studying philosophy, religion, biology, music, and other things that do not necessarily contribute to a fully optimized life.

I too have found a wonderful community in the maker scene, and it has fed my creativity. This can certainly be a good place to spend time.

> For me it's too much of a coincidence that the two most generally intelligent systems (ANNs and BNNs) are both "linear networks of activation functions".

I am unconvinced that "linear network" appropriately defines BNNs. Can you clarify this?

This is not wrong, but also not entirely right. There is a model called T0pp (T0 plus plus) which was fine tuned on simple logic problems, and it is capable of solving novel logic problems. This implies to me that there is more here than we've discovered.

Additionally, the whole point of fine tuning LLMs is to give them domain-specific knowledge. If you couple this with search/QA capabilities, the results can be quite impressive. I've not seen them in the wild yet, but I've played with them myself, and the performance is surprisingly good.

This is why we've built security policies at Mantium. You can run the input and output through an offensive speech detector, and halt replies the prompt if "badness" is detected. This is, of course, an imperfect system because philosophies around what is offensive can be very diverse, but we find that security policies are helpful.

UPDATE: These got a fair amount of traction, and I removed them out of an abundance of caution around deployment regulations that OpenAI enforces. Also cost considerations. I don't want to hijack the thread away from OpenAI, but you can also build stuff with Cohere and AI21 on Mantium, AI21's J1-Jumbo has pretty good performance, and Cohere just put out some significant updates for their models.

UPDATE 2: I couldn't help myself. I think this stuff is pretty fun. So here's a biology blog post generator using 2 chained Cohere prompts :) https://11292388-8f03-42d2-8a68-7039b24fcc2e-i.share.mantium...

This is, in my opinion, one of the biggest challenges with generative models right now. I'm not sure if this is the industry-adopted term, but I call them hallucinations. This is why I don't just pipe it straight into my blog, but rather use it as inspiration for a blog post that I write myself. It is easier for me to edit and expand on something that is already written, though.