They say AI will destroy humanity because they believe it. OpenAI and Anthropic were created by people who believed this. There's nothing nefarious about them saying this.
Why are they still building it? Because each team thinks that THEY are the ones who can prevent it from destroying humanity, but they have to get to AGI first, before the other teams make an AI that does destroy humanity.
But also, if AGI doesn't destroy humanity, it would be the most powerful weapon in the world, and they want to be the ones in control of it. Keeping the focus on Armageddon distracts from the real and severe problems that arise if a single person, or even a small group, controls an AGI.
It makes sense that this is true. We've been systematically underbuilding housing since at least the 1980s. Lots of places had old, run-down surpluses that were being "revitalized", hiding the issue for decades. But now the housing shortage is being felt almost everywhere in the country.
And yet, every year, we're still underbuilding relative to our population growth, thus making the problem worse. Unless our population starts to shrink like in Japan or Russia, we're either going to need to build A LOT of new housing or deal with an ever increasing population of homeless people.
Even a slowly shrinking population is probably not enough to fix the problem, because rich people like to buy vacation homes, and we have a lot of rich people. We'd need rapid population decline, like what would happen from some catastrophe. I recommend building homes instead of that.
The problem of homelessness isn't due to a lack of money. The problem is people don't want housing for the homeless (or any housing at all, usually) built near them. When there's a shortage of housing, there's gonna be homeless people, it's basic logic.
The amount of money required to fix the housing shortage is negligible for a country as rich as the US. But money isn't the problem here, politics is. And no amount of money can fix that.
I've been programming since I was 7, and I've always viewed software development as a means to an end. The alternative seems crazy to me - coding without a purpose? Why on earth would you want to do that???
I've been very successful in my career. Things I've built: Bloomberg's domain-specific language and simulation engine for asset-backed securities, a custom database that can process 100,000s of writes / second and 10,000s of reads / second, and the robotics framework powering the Cruise self-driving car (RIP). I retired at the age of 33.
The domain you're working in is usually more important to fully understand than software engineering concepts, although I try to understand both. But I don't really care about software development for its own sake, and I welcome LLMs replacing the more annoying grunt-work parts of the job.
Why are you discounting real wealth? I am a millionaire. If I spent my millions on buying a modest house in Palo Alto, does that mean I am no longer a millionaire? Have I lost all my wealth just by purchasing a home?
Any increase in the number of homes reduces the homeless population. There are so many homeless people in CA mostly because there just aren't enough homes for everyone there. It won't completely solve homelessness - many people living on the streets now need social services and therapeutic help before they can afford any rent again, but there's plenty of homeless people in California with fulltime jobs paying like $30k/year (around the median salary in France). If you build enough homes, these people would be able to find housing, and social services would be less strained for the people who really do need help.
Honestly, if you want a truly progressive city government in California, you probably need it to be 100% private. If it's public, the regressive California constitution comes into effect, and now you can't make property taxes be high enough to pay for city services, you can't increase the property taxes on long-term landowners, and those low tax rates get inherited by the property owners' heirs. You have to pay for city services with income/sales taxes. It's basically feudalism.
With a 100% private development, you can have a land value tax - it's legal if you just call it rent, and you can increase it however much you want (at most it's capped to 5% + inflation, a lot higher than the 2% (not inflation adjusted) cap for property tax increases).
And no, this won't be social housing. NIMBYs will say they want social housing, but they won't vote in taxes to pay for it (in CA, every tax increase must pass in a ballot referendum), nor do they actually want it built anywhere, either. Because only private money is being used for this development, it will mostly be market-rate housing, and that's fine. Or most likely, it will be nothing, since rural NIMBYs will block it.
Reading the comments here and elsewhere about this project is so vindicating for me.
In cities, NIMBYs will say "why should we build anything here, just move away and build your dream city somewhere else". I knew that there was no way that "somewhere else" would welcome construction of a new city, because there's people living out in the boonies everywhere. Those people moved to the middle of nowhere because they want to be far away from others - they're the last people who'd support new construction nearby.
So here's a new city proposal, paid for by private money, that won't take away anything from anyone, and even people who live far from the area seem to oppose it, just because it changes things.
Where are people supposed to live? There are not enough homes in cities, not enough homes in suburbs, and rural areas don't want new construction either. So where are the new homes for a growing population supposed to go? Or do y'all just want to keep increasing the homeless population indefinitely?
No, you won't be blocking the noise of an ambulance with noise-canceling headphones, that's not how noise-cancelling technology works. They only block persistent sounds, which an ambulance siren isn't as it changes pitch rapidly.
Does this fundamentally disagree with anything I wrote? Before Tumblr became a place where some people showed off their inscrutable social justice terminology, it was mostly a place where hipsters showed off their inscrutably cool aesthetics.
The ever-changing euphemistic language seems to me like mostly just a new way to be a hipster. The new vocabulary is a shibboleth to distinguish people who got the right sort of education from the right sort of places. Your knowledge of all the proper euphemisms to use sets you apart from the uneducated rubes.
The problem is, it sets you apart from the people you're allegedly trying to help, given that the marginalized are rarely those who got an elite education at an elite institution. This is how Republicans, a coalition of rich tax-avoiders and poorer people resenting the contempt of the elites, stay in power. And their political program - slashing benefits, cutting social services - hurts marginalized people the most.
By creating a way to be a "good person" that only the most hip and educated people can follow, the language policers are creating a rift between themselves and the people they are trying to help, preventing a political coalition from forming that would be able to pass helpful policies.
He has already "suffered" the consequences. Google stock is down ~35% from its highs, so his compensation (which is mostly stock) has already been reduced by that amount. Following the layoff announcement, it is up by ~3% in premarket trading, so he is being rewarded for laying off these people. This is how the system works, designed entirely for the benefit of holders of capital.
https://sfzoning.deapthoughts.com. It changed the conversation about zoning in San Francisco, and lots of politicians and activists now quote the topline stat (apartments are illegal to build in 3/4 of San Francisco).
Wage theft is committed by millions of employers, this "theft" was committed by just a few people. While I agree with your general sentiment, the resources required to prosecute wage theft would be vastly greater than to prosecute this case.
That being said, they should definitely prosecute wage theft much more than they are now. Maybe the fear of jail time if wage theft were actually prosecuted a few times would deter most other potential wage thieves.
It's funny remembering how people would say stuff like "if you aren't paying for the product, you ARE the product", and someone even tried to make a Twitter clone where the only differentiator was that you'd pay for it to "prevent" ads and tracking.
I thought back then that even if it succeeded in gaining enough users to beat Twitter, it would fail because eventually investors' demand for growth would force them to add ads anyway. You see this happening with cable TV, Roku, and now even Apple.
The only way to prevent this is to have a company motivated by something other than growing profits, but that comes with its own problems: stagnation, bloated bureaucracy, and capture by special interests without profit to keep the organization honest and lean.
I don't know what the right solution is, only that in very competitive for-profit markets, companies don't dare to worsen the user experience with ads lest they lose customers to a competitor. Given the barriers to entry for developing a new smartphone, including the strong network effect of the app store and OS APIs, this is not that type of market. Most androids already have ads and bloatware, so Apple isn't facing any competitive pressure on this front, sadly.
His argument makes no sense. If this is indeed why they are banning people, why keep the reasoning a secret? Honestly, every ban should come with a public explanation from the network, in order to deter similar behavior. The way things are right now, it's unclear if, when, and for what reason someone will be banned. People get banned all the time with little explanation or explanations that make no sense or are inconsistent. There is no guidance from Twitter on what behavior or content or whatever will get you banned. Why is some rando who never worked at Twitter explaining why Twitter bans users?
And how does Yishan know why Twitter bans people? And why should we trust that he knows? As far as I can tell, bans are almost completely random because they are enacted by random low-wage contract workers in a foreign country with a weak grasp of English and a poor understanding of Twitter's content policy (if there even is one).
Unlike what Yishan claims, it doesn't seem to me like Twitter cares at all about how pleasant an experience using Twitter is, only that its users remain addicted to outrage and calling-out others, which is why most Twitter power-users refer to it as a "hellsite".
The internet is not a public square for the same reason the world isn't a public square - it requires effort to see things from outside your own community. Twitter (and Facebook), on the other hand, will regularly push things to me from fascists, communists, "environmentalist" degrowthers, and other weirdos who I have no interest in hearing from.
Why are they still building it? Because each team thinks that THEY are the ones who can prevent it from destroying humanity, but they have to get to AGI first, before the other teams make an AI that does destroy humanity.
But also, if AGI doesn't destroy humanity, it would be the most powerful weapon in the world, and they want to be the ones in control of it. Keeping the focus on Armageddon distracts from the real and severe problems that arise if a single person, or even a small group, controls an AGI.