salmonfamine

Fair points —- I agree that the article is actually quite well researched and to-the-point, once you get past the title.

I guess I just see this exact argument being made all over the place, so I’m a little weary of it. But this article is one of the best iterations of it that I’ve seen.

I think this is a conversation worth having, but this is a tech news board. This isn't exactly on topic, and is guaranteed to be inflammatory.

Why was this flagged?

> From its training data GPT-7 might notice

> But its "aligned" so might understand

> Using this information it decides to hack

I think you're anthropomorphizing LLM's too much here. If we assume that there's a AGI-esque AI, then of course we should be worried about an AGI-esque AI. But I see no reason to think that's the case.

That's an interesting perspective.

It seems like it's trained on entire soundwaves. I'm curious if you'd get a better result by training it on transcribed MIDI and then taking the output MIDI and plugging it into VST's.

Seems like you would still get that "central brain" compositional approach without the garbled sound quality and unidentifiable instrument noises.

They are just very bland and a little musically incoherent. They pass the basic sniff test of, "this sounds like music" but the more you listen the more they fall apart. It's a good muzak generator, but that's about it.

I wonder if there's some sort of game-theory esque problem with cars.

As in, the more people drive to a destination, the less pleasant and convenient it is to walk the same route, and the more attractive driving becomes.

I think, also, the more driving you do in general, the more it feels like the default.

The average American is greatly affected by SV in many ways, hardly any of them positively.

I don't really have to propose solutions, actually. My point was the problem, and its inconvenience for many assumptions many Americans have about housing, cities, markets. I don't have to have a ready-made policy prescription on hand to recognize that.

Here are some potential solutions anyways.

Some of these would be unpopular, but we could consider;

1. National standardization of zoning (like Japan) 2. Federal funding for local rail instead of highways (this would probably be the biggest one) 3. Municipal incentives for densification 4. Federal laws regulating the lobbying power of neighborhood associations 5. Beheading local gentry at the guillotine before inevitably being placed in its open maw ourselves (vive la révolution!)

Yeah, Austin and Denver et al have also failed in their housing policy, for pretty much the same reasons as SV. No argument there. But it's hard to deny that SV's wealth explosion and exodus has not contributed to the problems in these cities.

Yes, they should build more. But again, that's just pushing the core problem away from SF (one of the most notable cities in the US for over a century, with a history of economic booms and high migration) to what was until recently a large-ish college town known as a haven for slackers.

Having lived in one of those cities as a low-income resident, a medium-income resident, and now a high-income resident, I assure you that the high income residents do much less for these cities than any other group. And no, taxes do not make up for it.

Yes they do. But we -- that is, our nominally democratic government, a "power structure" which you have arbitrarily equated with a collection of private corporations -- have placed restrictions on immigration, and some light restrictions on foreign ownership of real estate itself, for precisely this reason.

Secondly, Canada, whose real estate markets have been even more seriously disrupted by authoritarian regimes across the Pacific Ocean, have already placed outright, albeit temporary, bans on citizens of those regimes from parking their capital in the domestic real estate market.

Your characterization of my argument is specious and straw-man-esque. You are assuming that globalization is a given, that borders are always porous, that neoliberalism is the only way of looking at the world. I am kindly inviting you to reconsider your point of view, in light of evidence that it has made housing more unaffordable for the many Americans, disrupted their livelihoods, and dissolved their communities.

Or you can keep making colorless jokes to yourself about the property markets on Mars, up to you.

The discussion here regarding how fleeing SV workers is disrupting property markets in other cities in the US brings up an uncomfortable but, I think, unavoidable point: local housing policy has nation-wide consequences.

If SV and SF allowed for the construction of additional housing, there would not be a downstream housing crisis in Denver, Austin, etc. The policy decisions of a handful of city councils of the suburbs on the West coast has had a huge impact on these cities -- and has indirectly displaced residents, disrupted local economies, and pushed many people into homelessness.

Housing policy in major economic zones in the US is a national issue, and should be treated as such.

This doesn't even touch on the wider problem of income inequality itself.

If we're being completely honest, the average American would be much, much better off if SV had never produced a single unicorn.

Interesting, I pulled this from a very recent tweet from a climate scientist. This article delves into it in more detail: https://www.carbonbrief.org/explainer-will-global-warming-st...

Curious to hear your take on that.

This is not true.

https://climate.nasa.gov/climate_resources/305/graphic-when-...

There's obviously quite a bit of complexity here. Huge difference between being terminally ill and getting used to some bad conditions for a finite period.

Mostly just this:

> I DO expect that short-haul public transportation would suffer a significant decline in ridership, as it would be outcompeted by driverless cars.

My contention here is “outcompeted.” Ultimately, this is a policy decision —- at least in the United States. In regions where there is both existing rail and highways, there might be good conditions for a competitive market to exist between the two. But, that competition will also rely on energy pricing and EV availability, as well as policy decisions. As you mentioned, density is a huge factor. However, there’s not sufficient evidence that self-driving cars can solve the congestion that comes with low-density.

In short, there’s far too many factors to make a big prediction about the fate of either at large.

In general, my contention is with those who tout self-driving as a transformative technology that will turn our congested highways into high-speed infra, and all with a higher degree of safety than existing cars. We don’t know any of that.

We do know that short-haul public transportation can work very well.

There is no reason that self-driving will create a transit modality that is more conducive to smaller vehicles. That is an energy cost “problem.”

Ok.

As opposed to highways, which magically have none of these problems.

That only raises more questions. People will buy them, of course. People buy cigarettes, heroin, and NFT's. Will they actually meet high enough safety standards in these places? Will they reduce commute times?

In other words, are they actually going to deliver the long-haul, end-all-be-all transit solution that many of its proponents imagine they will? Or is it just a nice-to-have feature that will make driving a little safer and a little easier for some people? And if so, is that value really worth the incredible amount of effort that's gone into making it a reality?

Well, that's the problem. Self-driving cars does not solve that problem.

Walk.

EDIT: or just drive, as you do now.

Self-driving cars are a solution for a problem that has already been solved. Build trains. But of course, this doesn't benefit automakers, requires a move towards high-density housing that NIMBY's and certain classes of investors oppose, so instead we'll create an entirely new class of problems by allowing these companies to unleash their insufficient technology on public roads.

Then we agree -- both false.

You're drawing a huge false equivalence here. The idea that we can deploy renewables at scale within even two or three decades sufficient to reach net zero is patently false.

1. There is no "10 year" time limit. There is a window of time to keep our overall warming below 1.5c, which is virtually impossible regardless of which policies we choose to pursue, collectively.

2. There is no indication that we will be able to produce lithium-ion batteries with a high enough wh/kg density to create high enough energy storage capacity to replace carbon-based energy sources. Until such a time as that technology becomes possible and production-ready, we will need other sources of energy.

Not building nuclear with the assumption that grid-scale battery storage will become feasible is a potentially catastrophic policy decision.

That is also not true, and not supported by the IPCC.

What do you mean "we can't wait that long?" Any time we start reducing emissions is better than never. There will never be a "too late" point to reach net zero.

Secondly, what is the alternative to reducing our carbon outprint in 10 years?

[citation needed]

You just have to practice. I used to be quite bad at it, but I've improved greatly.

Here's what I recommend: start with a specific topic -- maybe HashMaps, Graph Problems, Dynamic Programming, Greedy Algorithms, whatever. Start with the easiest problem on Leetcode for that category and work your way up to a "Hard" problem in the same category. It helps build a mental "muscle" for that type of problem, and it will be easier to activate that "muscle memory" whenever you have to brush up on Leetcode again for your next round of interviews in the future.