This. The lower end of those "upper middle class" numbers may qualify for welfare in some tech hub cities.

They do point out that it varies by location, but really their number range is terrible. "classified a family of three earning $133,000 to $400,000 in 2024 dollars as upper middle class." From the HUD Section 8 income limits, expensive places the lower end of that is considered low income, like San Jose 143,600 qualifies for Section 8, versus cities like Akron where 72,250 is low enough to qualify. Location, location, location.

As this is /. lots of us live in tech hubs that even though we don't like the costs, they're very expensive places to live. In my current city despite being a full hour commute from the city center 130K is still solidly middle class. Not poverty, but not upper crust either. That income wouldn't require a trailer park, but would have a hard time affording a 3 bedroom / 2 bath home (they'd add another 45+ minutes to the commute distance), one or possibly two small vacations per year.

In tech hubs especially, those household incomes can be very middle class, not upper-middle, and in some places, lower class lifestyles.

Statistically it is not the severity of the punishment, it's the likelihood of being caught and facing consequences.

When the likelihood of facing consequences is high, even frequent offenders comply even when consequence are minimal. Even something like line jumping, it's about if they think they can get away with it, not the seriousness of consequences.

Traffic violations are similar. When "everybody knows" a curve holds a speed trap people drop to the speed limit, when they are past it they lay on the accelerator. When I see a bunch of break lights ahead frequently I will hear GPS call out "speed trap ahead" right where the others started slowing. The perceived risk of enforcement, not severity, is the biggest factor.

Scammers know the risk is minimal. In India many of the big operators include people in politics and police, it doesn't take much of a cut to bribe officials.

In general they've been on the right trend of only playing after interactions.

Most of the time they're ads, and even if they do manage to slip through my ad blocks and DNS fliters, I still don't want them playing.

Waste of bandwidth.

It's one of those issues that often are easily ignored until suddenly it's game-breaking.

For many games that stream assets and build shaders on the fly, if there's a bit of blurriness and stuttering when you first enter an area there are many players who can forgive that. Having that same experience walking into a boss's lair and suddenly the game is choking and stuttering as resources are processed, that's a fatal flaw that can make it difficult to play, or even outright kill the player while loading.

The difficulty is that usually it's an all-or-nothing experience, a tradeoff, games can choose one or the other. There are plenty of games that do build them at runtime, and others that have the long slow progress bar at startup. Different choices give different experiences. As annoying as waiting a couple minutes at program startup is, having a cutscene or a boss fight stuttering for shader compiles is far worse.

BOOK:
The Mindful Way through Depression: Freeing Yourself from Chronic Unhappiness
https://www.amazon.com/Mindful...

No drug necessary.

The Underlying issue is: Why are some people depressed?

If we understand that, it is very likely that no drug will be necessary. People won't become depressed.

Windows is crashing because users are doing things that cause crashes.

This issue needs to be investigated in MUCH more detail.

I don't have crashes.

Sadly while your logic sounds spot on, the reality is, that TV still does ACR (basically hashing each screen sometimes multiple times a second) and builds one hell of an accurate profile of everything you watch. Even if the source is HDMI. And it's still sold.

Food for thought, that HDMI cable to your dedicated device, it likely supports networking, so the TV has a path to the internet even if you don't connect an ethernet cable directly to it, or add in your wifi creds.

Grok was constantly say it was doing something that it had ZERO ability to, and I kept calling it out and it kept apologizing and then immediately doing it again.

As a guy who spend 5 figures a year on Ai, the last thing I want is that. I know Claude and ChatGPT also do it, but Grok was doing it CONSTANTLY.

I've heard of traffic 'accidents' where a diabetic went hypoglycemic, passed out, and drove into other cars, pedestrians, etc. For some reason the news always reports it as a 'medical event', but the point is, people are killed by diabetics due to their diabetes way too often. It is an apt analogy. An idiot decides to drink and drive an puts people at risk. An idiot with diabetes fails to control their blood sugar and decides to drive, putting people at risk.

Yeah, it's easy to add more code to fix stuff that should be deleted. Just have the launcher code call your new code which bypasses all the old code. The old code can happily remain, it'll just never be called. No special skills required. If you look at the windows codebase, you'll see this technique everywhere.

While that's the hype, it's not going to be the reality.

Yes, what they are doing has a market.

Yes, it will absolutely allow many of the masses to do what programmers have been able to do for ages. It will change the market, the cheese will move, but it won't destroy the marketplace for programmers.

I think work in graphics design is probably the best parallel. People freaked out in the 1980s when home computers could make banners and flyers. As the software advanced, you got more and more people doing Word Art, and enormous clipart catalogs let office secretaries make good looking office flyers, creative garage sale fliers, church bingo night announcements, and much more. LLMs let people continue to create this type of thing, and print-on-demand services let them send their creations out to make custom stickers and such. But most critically, NONE OF THOSE PEOPLE were hiring graphics designers for those jobs before. It enabled the masses to do some of what graphics designers do, but when it comes to real ad campaigns and professional marketing, companies know investing a few hundred dollars will bring in a few hundred people from the community, investing many thousands or millions are essential for large regional or national campaigns, those jobs continue to get the professionals.

More people making vibe-coded websites that satisfy their specific needs? Great. They weren't hiring a team of programmers for software development before, and they're not hiring a team of programmers after. Executives that claim they'll cut costs by 90% by firing all the professional programmers are in the hype, they either don't understand the work being done or are playing the field. They may do well in their quarterly financial statements, but a couple years down the line the company won't have anything of value remaining. The CEO will be long gone, sold his options, collected his golden parachute, and moved on to the next company to be restructured. Investors will have gutted and sold everything of value from the company by that point as well, they'll take the hype bubble, milk it, then dump what remains in an asset fire sale. The companies that continue making great things and not seeking the bubbles will continue to create good value, leveraging the tools where appropriate but still hiring skilled workers to create products with lasting value.

There will always be changes in who is the winner and who is the winner this quarter. Certainly plenty of profit-seeking investors care only about those quarterly results, not the products and services on offer. There are companies that will grow and companies that will die, nothing new is there. It's good that more people will be able to have more custom program options, just like WordArt and clipart collections allowed people to easily make their own fliers. Those who want a specific vision in marketing can start with "here's my interpretation from an image generator, but I want it done better." Similarly when a small business needs a team of developers to build a program, the customer can also bring in what tried and failed and what they want to see differently, and they come back with a better bid being able to reference what the client generated using AI as a starting reference for building the professionally-built items.

Agreed. The ESP32 family of processors are popular for good reason. Some of the chips support RISC-V.

Unfortunately there is far more to the question than the Ask Reddit post contains, cost, processing speed, memory requirements, software needs, and much more. Even so, the ESP32 family and earlier ESP8266 have been popular in IoT devices from smart lightbulbs, watches, cameras, and even light industrial use to Arduino ecosystem and student devices for over a decade now.