pattt

I actually agree with most of your policy list. But I think the reply misses my point. I'm not saying efficiency gives worse results than the status quo. I'm saying it can't produce any results until you've made moral choices first. For instance if the goal is maximizing long-term growth per dollar, why not cut retiree benefits and disability support? Saying we shouldn't is a moral judgment, not an economic one.

That’s true. You can’t just build the government based on economic efficiency.

I don't get these "distilled" comments. How would one realistically go about distilling the 1-2TB closed source frontier model behind the oracle (provider) ? Just genuinely curious.

How in the world was it possible “back then” to be considered innovative and cutting-edge with just so few simple but beautiful products (iPod Classic, iPhone 4..) ?

It seems like Alzheimer’s and prion diseases like Kuru share a core molecular mechanism with misfolded proteins propagating and causing brain lesions. I find it really interesting that the gene protecting against Kuru has evolved naturally when according to Wikipedia, prion diseases don’t seem to have any effective treatments.

I understand they have been criticized because they play capitalism particularly well, but there are many good points in the original article and it didn’t sound like an outright PR.

It would be interesting how many iterations of LLM-based PR reviews and fixes it takes to turn the code into horror. It’s already started happening.

Do we have any solid evidence these models can outperform Western models in terms of quality? Or is it more: because they are forbidden, they can't get enough training data, visibility etc. to compete?

Is there a way out of this though? The seed is already planted and everything else is just a PR.

Different regions play complementary roles in the global economy. The oversimplified description is that the US drives innovation and risk-taking, Europe prioritizes social welfare and sustainability and Asia powers manufacturing. Forcing everyone to compete in the US-style growth race would be disastrous - imagine the burnouts, resource drain and market instability if every country tried to be Silicon Valley. Plus, many Europeans simply don't share the "get rich quick" mentality - they value work-life balance and social security more. Whether we're being pushed into this growth-obsessed game is another debate. Ideally, regions should focus on their strengths while maintaining independence. Europe can keep its social model while smartly adopting tech, Asia can focus on sustainable industry, and the US can lead in innovation. It's like an ecosystem - diversity makes it stronger. The goal shouldn't be for everyone to play the same exhausting game, but to maintain different approaches that complement each other.

If you consider complex forms of life serving as an entropy-increasing phenomenon, then you might as well consider that the evolutionary algorithm is governed by such goals. It's even plausible to take the human behaviors like increasing consumerism and growth-orientation as being connected to this fundamental thermodynamic drive. Perhaps we'll find even more efficient principles to drive our consumption further. <extending on your rant obviously />

I tried taking notes for learning and brainstorming over the years but found it incredibly difficult, not because of a lack of note taking frameworks, apps or technologies but because _sustaining_ this process is hard. Reading posts like these makes me quite envious of author’s dedication. Recently I was contemplating about a possible lightweight middle ground solution where I think a more polished software would certainly help - _highlighting_. Imagine if you could simply highlight/select any text/image on the screen of any app (ok, let’s start with the few), optionally assign different colors, tags and have those highlights automatically synchronized in a sort of a searchable personal diary format. The main feature of such tool would be cropping and indexing what’s already read and seen as opposed to having to summarize/rephrase or enrich it with your own notes. Why I think software in this case reducing friction to a minimum would help? As opposed to taking notes, identifying illuminating corner-stone paragraphs and sentences mostly feels like an implicit process that happens naturally at least in my experience.

GPT3, Stable Diffusion etc. will really kick off. People will start accepting these blackbox Big Data sequence prediction models as a new norm, a powerful assistance for all sorts of domains. There will be some resistance but the argument that one doesn’t need to understand digestion in order to eat food and survive will eventually settle and the path of these models becoming a natural extension of our minds will be laid out. We will see different interfaces (speech recognition, OCR) and pipelines being created to fit these models into real problem solving pipelines and address their inaccuracies, i.e., with additional fact/spec checking and computation logic. some successful startups and products will emerge mostly from folks who already started their ventures into this space.

Fully agreed, the way I'd approach this would be that the said mathematical proofs about certain truths "by definition" rely on human logic as the main building block and substrate. Logic is a human basis of agreeing which seems necessary evolutionary. Counting and separating observable objects turned out to be quite necessary for survival as well. Hence this statement seems to imply that aliens would need to have a corresponding logic reasoning system and observational abilities. If that was the case perhaps there would be a strong inclination to believe that the isomorphic reasoning would be deduced.