zmnd

It’s easy to dismiss a problem by just saying that someone is just an “older person” who “didn’t adapt”.

Not my quote, just pointing out that not everything can be solved with money. But yes, I think it got worse over time.

How do you fix the mentioned problem of "filled with unpleasant people" with money?

If a doctor diagnosed me with the same result every time I go to another room, yes I would.

At one of the FAANGs I work, it's a relatively smart autocomplete. Saw some examples of autogenerated e2e tests. I think it's quite decent as a language tutor.

But besides that, agree, not many good examples, and it's been a year since gpt-4 release.

What you describe sounds like an immigrant, not an expat.

I am originally from one of the poorest countries, and my family constantly struggled with food for about 10 years. So I am well aware of the struggle. As well as I am exposed to poor people who immigrated to the US.

With all of that, no, I don't believe that there are people who actually starve in here. There are extensive food bank networks, generous religious organizations (most of them not that pushy), and in general people in the US are kind, despite what media says. I saw plenty of videos of people from my country showing food they got from a food bank, and most of them highlight how generous it is.

Despite all the struggle, my family never stole. And poor people around me rarely stole as well. Thieves whenever caught were judged by people around, and there was always social pressure to keep everything orderly.

What surprised my in the US is how normalized the whole thing is. It's always "stores have an insurance", "they are just trying to survive", and other excuses from people who never lived in poverty and just want to feel better about themselves.

Can you give an example of what question someone asked GPT-4 and was misled? And how that question was better answered by one of your tutorials?

How do we distinguish if it's "beta" or not? Waymo didn't kill anyone, is it beta? How does non-beta rollout look like?

But it's still more (or at least fairly similar to me). Far from "old American disease of not ensuring competition".

I shared the whole quote, but the goal was to just highlight that the number of banks is fairly similar.

I agree that the number of people is a better characteristic in this context, however the number of countries is important. EU has many different cultures, languages, regulations, so they have to have more banks. But looking at data, despite these considerations, EU still has worse competition. So still not sure what exactly parent meant.

The US is home to over 4,700 FDIC-insured banks. It’s not far off from the EU, which has 5,171, but stop and consider that the EU consists of 27 countries. The UK currently has 365 banks, and Canada has 83

You are saying like it's any different in Europe. Dassault Systèmes CEO Bernard Charlès makes €20.6m with company revenue of $6.05 billion. For reference, Google CEO made $226 million (~10x) with company revenue of $279.8 billions (~46x).

Without taking into account compensation difference (that is also significant), almost no other developed country has such favorable tax regime (again ignoring temporary programs). Check an average FAANG salary on levelsfyi for seniors and compare how much would someone pay taxes on it in Germany.

Now calculate the difference in income taxes on 200k+. You are just paying it differently.

Mind to share the problem?

Why would they optimize it if the service is provided for free?

Why would they buy them out? The main value prop for the client is no ads, reddit wants to serve ads or provide a subscription. So while it was great, reddit provided free infra, that Apollo resold to their users, it's fair to expect it to end. 20M is something like 15-20 dollars per year per user for them, seems fair.

But at the same time, groups who are responsible to resolve the homeless issue quote that only 8% of people are out of state based on HIRD research. So what is it?

The internet is big and I just didn't look for it very long.

You'd think so, but for the majority of homeless programs I was interested in, I couldn't find any data. For example, I was trying to find efficacy of tiny villages, what are their goals, how they track graduation rate. What's even their success criteria? Nothing, despite my decent google-fu.

You'd think so, but policy is often penny wise and pound foolish. I think the biggest obstacle is that "free money" just doesn't fit in our culture very well (or maybe it just doesn't fit human nature).

For policies, sure. What about non-profits? Why wouldn't someone who believes in the efficacy of such methodology organize something like that? Changing someone's life for the better with only 3k sounds like a no-brainer to me. I would personally contribute financially and direct other people to contribute.

If there were a few studies, and it's overwhelmingly successful, why do you think it's difficult to find the results? SF spends something like 90k per homeless person, conducting an experiment with 3k per person sounds like a drop in a bucket. I bet, showing great results on a group of 100 people, would get more investments into this experiment.

Because again, that's their incentive. They use whatever ads they get paid more for. Worst case, they are going to lose frustrated users like you, and make much more on others. Products that want to bring a better user experience, just can't survive, because even well paid users can't justify a couple of dollars for their service. For games there's Apple Arcade, for news - Apple News. They have some decent alternatives. Support great products that you like.

Sure it does, there are definitely exceptions and Angry Birds is among them. And your example is probably from a while back, since you can disable tracking on iOS from 2019 or so.

I intentionally find products where I can pay. I like YouTube, and I gladly pay for it to not see ads. It's a simple transaction, they provide a service - I pay for it. Sure, you can complain about sponsored content and other crap, but overall YouTube premium experience is great. Whenever I want something to play on my phone - I find a cheap game and just spend 2-5 dollars. In some of the free games, I intentionally spend money on cosmetics to support developers. Without this relationship, you are just a non-paying user who brings no value, and you have misaligned incentives with the app developers.

So I don't understand the mental gymnastics of well paid people (given the audience of hackernews) for why they don't want to pay a few dollars. This way you would support people who provide value to you and dictate how the industry should evolve. Instead, we have a Guac-A-Mole game with advertisers and complain every other day on forums.

If you don't want to see ads, why you didn't want to pay a couple of dollars for an app? Or how do you see developers compensated in this case?

like.. apple arcade?

To me, it prefixed with just "As an AI, I do not have opinions or favorites. However, I can share with you three notable and commonly appreciated Power Cards from the game "Spirit Island", as it existed until my training data cut-off in September 2021. Remember that the "best" cards can often depend on the specific circumstances in the game, as well as the particular strategy and Spirit you're playing.". But then just shared the cards, nothing about that it was not sure about details. Card selection was decent, but details like resources, powers, and so on were off. But all sounded realistic. Shared an example below if you care.

Copy-pasted exact query you shared. It still provided information that I thought was correct, but after further inspection just sounded correct. It's GPT-4 without plugins.

What are top 3 Spirit Island cards and provide their descriptions? Only provide information you are certain about. If you don't know the answer, say "i don't know".

2. "Vigor of the Breaking Dawn" (Major Power Card): This card costs 6 energy and is a slow power card. It allows you to push all explorers and towns from one of your lands. If you have 2 sun and 2 fire elements, you can instead replace each town with an explorer before the push. If you have 3 sun and 2 fire, you may also do 2 damage in each land where you push one or more pieces.

https://spiritislandwiki.com/index.php?title=Vigor_of_the_Br... It doesn't cost 6, it costs 3. And it's not slow, it's fast. Incorrect required resources. You can deal 2 damage, but for pushed Dahanies, not for pushed explorers. As I said, all sounds convincing, but wrong. More common queries are usually fairly good.

I think the main reason for division is that everyone projects to their own use cases. I have been using gpt-4 for quite some time and also couldn't understand why someone would say that it just produces something that sounds like a real answer. But then I found some queries that can definitely be described as "sounding like truth". So your personal experience probably wasn't what was their experience.

For those curious, I was asking gpt-4 about the top 3 cards from my favorite board game, Spirit Island. All three of them sounded really convincing, having the same structure and the same writing style, but unfortunately none of them existed. So everything that fails outside of most common use cases would probably have an experience of convincing hallucinations.

We have tried neural nets before as well. But technologies and constraints are changing. I just took a screenshot of the Facebook feed, and segment-anything model was able to identify most of the elements, even without proper fine-tuning. Scrapping billions of elements with classifications is also something fairly trivial.

I, perhaps naively, think that training a better model(s) that helps with accessibility is a much better path forward for users that need it. Just a genuine question, not sure why I am getting downvoted. By no means, I am implying that this is not important.

For people with more accessibility expertise, why instead of pushing all websites to be accessible, there's no push for a better accessibility AI? I would imagine that having a better interpretation model, implemented by a couple of tech giants, would be a more productive use of our time than ensuring that all aria labels are in place? Is it because nobody is working in this problem space? It sounds fairly straightforward problem to solve, comparing to some recent achievements.