tomohelix

The article to me is about the cost of a box to sleep in SF. The person they interviewed was just one of the renters. There are all sorts of people living in that place. Most I would wager are the people who cannot afford anything better.

Why do you think minimum wage workers should not be allowed to live in SF? If everyone is writing code, who would make the food? Who would deliver the packages? Who would clean the workspace in your office? They don't deserve a place to live near you even though they are providing you with all kind of services? Genuinely asking here because I can't fathom what you were saying when you said SF is not for minimum wage workers.

The apathy is what I am angry about. The sentiment that "this is fine" is what allowed this to happen. Laws and regulations need to be adjusted to enable better housing for more people. But they get blocked because those who hold the means don't see any problem with people below them living in squalid conditions like this.

Make no mistake, I see this as a terrible living situation. Just because some war time refugees 100 years ago lived worse doesn't mean one of the richest cities in the world should have the same. It is a terrible and dishonest comparison.

Someone working full time should at least be compensated enough to have a proper living condition.

A box is not a proper living condition. For a minimum wage worker, $700/month is basically at the limit of what they can afford for rent. And yet this is all they can get? And people are cheering for it?

Why do you think it is temporary? You don't seem to understand how much is $700 to a minimum wage worker. Even in SF, that is nearly 1/3 of their income. If we take the common requirement of 3x the rent then this is basically the most a minimum wage worker can afford. Sometimes, people can't do any better than that. Are you saying that even if a person work full time, they don't deserve a better living condition than a box?

This is eye opening to me how the housing situation is viewed by the SF upper class. I have always considered it a dystopian nightmare when the workers have to sleep in a coffin and work full time. It seems not many people in SF share my view.

I would rather have a tumor be cut open and removed than having a band aid put on it and left to fester.

If the housing situation in a place got so bad that a person has to pay nearly 1/3 of their monthly income just to get a closet to sleep in, it is not a livable place. A slum has a big mix of people but would you really consider it a great "metropolis" where you can have a life and raise a family?

Again, this is 1/3 of minimum wage for as much space as a closet. It is basically priced as high as it could to best exploit the poorest among us. Maybe it is a different in value between me and those who grew up in SF but I consider it inhumane to treat someone who is working full time like this. If you don't think so, let me ask you this: would you let your family, your spouse and your children, to live like this? In a box? Or would you do all you could to get them something better? Not everyone can do better than minimum wage, but it doesn't mean they deserve to live like this. That is what I mean.

There are solutions to the housing problem that is not this band aid on a festering wound. But it seems this kind of situation is perfectly fine to SF people. To which I have nothing else to say but amazement at how can a city priding itself on economic abundance and human right progress can also be utterly indifferent to the suffering of its fellows and even justifying it.

Looking through this thread, it is surprising that there are actually people who consider this "acceptable". Like it is not something out of a dystopian hellscape.

Amazing how much a person can be conditioned into accepting a terrible situation. And no, there are ways to fix this, there is just no will to do so, partly because of the people who see nothing wrong with it.

Is this something like short term vs long term memory? The context window for LLMs is its short term memory where you can tell it to do things or quickly define something and the LLM can learn very quickly even with just 1 example or a sentence. But it forgets immediately once the work is done. But for finetuning, it commits the knowledge into its weight network and have a "deeper" understanding? The cost is it takes more effort and energy to do so?

If so, let say in the future, we have an LLM with 100K token context windows but with a subsystem where it notices some knowledge keeps being repeated in the context and then store that knowledge for finetuning when the LLM is not doing inference. Basically a mirror of the way we human work? Is that possible? An LLM that constantly improved and can adapt to new knowledge?

Honestly, whoever convinced Microsoft to make Linux run in windows was a genius. I would have seriously considered a Mac had they not done this. It just opens so many possibilities in Windows for me and all of it is so easy and simple to setup.

Why is this flagged? I started to see a lot of good discussion topic get flagged recently. I thought HN is more matured than reddit...

The fun thing about biology is that if there can be something built by a true omniscient creator, then biology is the closest thing to it. All of reality and all of physics is used. It is simply what works no matter how arcane or difficult to understand the concept and mechanisms behind it is. Like all engineering projects there are still physical constrains of course but there is no limit on the technology allowed. There is simply no "cutting edge knowledge" or "unexplored science" to evolution. Literally everything can be applied within the physical constraints of a carbon-based lifeform

From what I remembered, string theory proposed that the extra dimensions beyond the spatial and chronological ones we experience are also compactified and technically hidden from our perception. In another word, they exist just not on a scale we can appreciate. If you accept that then the 2D plane with the 3rd dimension minimized to the point of irrelevance is not so different.

And I don't really care about what the bacteria "think". The point being there exists an extradimensional relationship between me and the bacteria right here in front of me. So at least to me, it isn't so outlandish that there can be something above this 3D plane and I am simply too primitive to understand it.

Of course, there is no sufficient proof right now so I do not accept it as a real answer. It is simply analogical reasoning. Also, I would not say it is unfalsifiable. Like mentioned earlier, the extra dimensions are simply tucked away and not easy to detect. They still exist and with sufficient technology, they can be probed.

The way I understand the extra "dimension" here is like me viewing a bacteria on a cover slide. To the bacteria, that 2D flat plane is all that matters. Sure, technically it is 3D, but the 3rd dimension is compressed so much it is basically nonexistent and the bacteria can't really be bothered anyway. They can't even comprehend the concept of dimensions.

But to me, I am fully in 3D. I am literally an extradimensional being to the bacteria. And I am so far above it (no pun intended) I am incomprehensible to anything that lives on that 2D slide.

Even if I assume the bacteria has some really good teachers and can build stuff (they actually do, they make biofilms that are pretty complex "superstructures"), the kind of thing they can do in their limited plane of existence is nothing compared to what a human can.

So while I don't believe anything that was released so far points conclusively to real aliens, I don't find the concept of extradimensional beings too far fetch.

AFAIK, there are two main issues. The first is ethics which is pretty much obvious. Experiments on human are not easy to get approved and near impossible for things like permanent changes to babies and fetuses. There are ethics boards on every company and university that regulate these stuff and they would almost immediately deny all request or proposal that approach this kind of work. You may get something going in less regulated countries like China but even then the international research community would not be kind to you and your career.

The second is theoretically, it still is not possible to "design" a baby. In biology, things are very different from pure engineering. The ways stuff gets discovered is like this: if you have a big clockwork machine, you increase signal to a gear, you notice the machine do X, you say that gear is associated with X function. Then you go to another gear, do it again, and try to map out everything in the machine and their function. Cool. Now the problem is you don't know interactions between gears. So you try to change 2 gears that both associated with X. Turn out when you do that, the result is not just X but XYD get changed. Then when you try 3 gears at the same time, turns out it changes XMA function. Worse, in another supposedly identical machine, changing gear X actually causes Y change because of a previously unknown gear that is a switch for gear X to affect either X or Y and that switch depends on some obscure environment parameters...

In short, biology is too complex and the way we are probing it is still too primitive and it takes a long time to even figure out what go where. We can try poking around to optimize one thing but it always causes unintended effects, mostly lethal. So the ethics boards are even less likely to approve a project like this.

Just as an example: CRISPR system. We know about nucleases and these enzymes and stuff for decades. But to make it something reliable enough to be used took so long this thing only became big recently. Biology is very slow and very tedious. It won't be like writing a software and wait 10min for it to compile. Experiments can take years to run and come out with nothing because a stupid mistake was made somewhere along the way.

Of course, I have no complaints or problems with things people do behind closed doors.

I only have issues with those who smoke so much it reeks even when they are not smoking or I am 10ft away from them and still smell it.

I believe this is a topic of sufficient intellectual weight for HN. There is certainly something wrong with the law enforcement system in the US and this is just such a blatant and clear example of it that we can point to and discuss about.

You do not have to engage if it makes you feel uncomfortable but these issues are real and very relevant to a significant number of people, including HN audience, and they deserve to be addressed.

Oh boys, crypto again.

I am honestly and sadly not sure why crypto is still a thing now. I am unaware of any significant or major product that came out of crypto except for all the "coins" and the NFT fad. I know that on paper, blockchain tech has great potential for security and bookkeeping stuff but I don't see anyone using it for that yet. Every time crypto comes up, it is always about some virtual currency that is notoriously difficult to transact and has extreme volatility.

Can someone explain to me why crypto is still a thing in 2023? Beside the "coins" reason?

In addition to software and computer tech in general are so easy to scale and automate, there is also the matter of responsibility and legal regulation.

Take the biotech industry for example. There are also plenty of acquisition and consolidation, but even the biggest top 5 biotech companies combined cannot match the size of one Apple or Microsoft. This is mostly because biotech is so much riskier and a failed product is extremely punishing. At any stage a drug can fail. Even after commercialization and release, a drug can be removed for plenty of reasons. One drug can cost hundreds of millions of dollars to get to market and years and years of R&D by hundreds of scientists going from discovery to production, validation, scaling up, and QC, and marketing. None of the steps can be skipped because the FDA breathes down their neck in every step. People will also sue the company to death if some fuckup happens.

Meanwhile, Microsoft can churn out a new software or product in a year from a team of 20 engineers. And if a fatal error happen, no biggies. Patch incoming and that is that. Lawsuits from bad codes aren't as severe as lawsuits from bad drugs. Leaks and exploits and all the "huge" issues get solved and the product continue to stay on market. A drug fuckup and it is very likely to be scrapped along with all the investments that went into it.

And so big tech continue to grow unchecked because they are just that: unchecked.

No offense, but isn't the NLP field effectively solved with the creation of LLMs, or at least for the majority of the tasks you would expect from an NLP application?

I am sure you can find some special areas or niches where traditional NLP approaches would outcompete a black box like LLMs. But with the LLMs becoming much more efficient now after quantization to the point you can run them locally, I think there is a good argument in saying simple NLP is basically solved.

Part of the reason I am still on Android is because I can install and uninstall whatever I want, including stuff like Revanced.

I still love to have the battery and the "it just work" reliability from Apple though. Let see how this play out in the US. If Apple doesn't hobble it too much I might get an iPhone for my next phone.

Guaranteed it will have some catches though. I also wonder how much it would affect the security side since Apple devices have been pretty secure so far with its walled garden approach.

RAM is easy. The hard part is making the unified memory SOC like Apple's. From what I know, Apple performance is almost magic. And whatever Apple is making, they are at peak capacity already and they can't make more even if they want to. Nobody else has a comparable technology. Apple is in its own league.

It is obviously the prison for an Elder One. The immense gravity is needed to keep them drowsy and not stirring too much.

Who need chains or tractor beams? A planet gravity is the perfect restraint one needs for these kinds of inmates. Low maintenance and high effectiveness. Let just hope they aren't getting any jailbreaking going on now that we have disturbed the quantum harmony of the prison by observing it...

Definitely a paid/hype article. People has been doing the "lots of automated generation" stuff for a long time. It won't work because if there is one thing our human brain is good at, it is pattern recognition. And no matter how advanced or detailed the generation is, we will see the pattern and get bored with it.

Even something as sophisticated as the LLMs get monotonous really quick after a while.

Aren't we running out of text to train AI? You need much more data to train a smaller LLM to match a larger LLM performance. Considering how everybody is starting to clamp down on their user's data, we might have issues getting enough to train something on the scale of PaLM to reach optimal performance.

On that note, everyone is probably salivating over Discord's data. So so much chat based and detailed conversations, all neatly organized into groups...

For a more esoteric and realistic scare, let me introduce you to Ethidium Bromide. The thing has as much toxicity as salt water with the amount they are used in a lab, but many places still treat it as high hazardous waste and demand specific procedures for handling and disposal.

Meanwhile, bleach is probably more toxic and carcinogenic than this thing but nobody say anything about special handling of pure concentrated bleach. Most scientists know EtBr is not too toxic but because of the administration and extra burdens of dealing with it to satisfy the rules made from outdated knowledge, they end up switching to something else.

Remember when Steve Jobs introduced the iPhone and its own appstore? And how developing mobile apps is now its own field?

A new interface is ripe for development. Once a user is used to such a clear and intuitive UI, they can't go back. And GPT technology is exactly that. ChatGPT is the herald of the new era UI.

It is not that high. Average houses are around 300-500k in Texas. That is like 5-6k in taxes a year.

Not high enough imo

While I understand the point of expanding the term to give representation, at this point it is a caricature. They tried so hard to look inclusive that the term become so unwieldy and obscure, nobody would use it except those who are so deep into it and those who want to mock it.

Also last time it was updated, there is an A in the term somewhere. Guess they dropped it or forgot it.

Buying one isn't hard. They are quite cheap. A completely broken house like the ones in the OP is probably begging to be sold. A 2k+ sqft livable house in the suburb of Tokyo is about $300k.

https://www.nytimes.com/2023/04/17/realestate/japan-empty-ho... https://archive.is/BDCnm

But these are traditional Japanese houses so quite a lot of renovation is needed to bring it up to a Westerner's standards. Still, all in all, if you are fine with living in a backwater village with barely any modern conveniences then 200-300k for a big house is not too bad.

Assuming you can speak Japanese and don't mind the xenophobia...

This is why I prefer anonymity for my online activities. You never know when or how your online posts will be used against you. Maybe everything I do now is appropriate and "moral" but in 20 years they would not and some nobody would grab those and slander me. You can never be sure.

Back to the article, I guess one good thing coming out of this is that the guy can apply for political asylum green card now...

Based on this article, I don't think oversupply is the problem here.

Also, the US economy is already overly dependent on foreign labor. From offshoring jobs to manufacturing to onshore labor, every sector of the economy is filled with foreign labor. How many things you own were made outside of the US? How many of your coworkers are immigrants? How many foreigners do you interact with when working on a large project at work?