mgrunwald_
mgrunwald_ on Hacker News
You have all the info you need in the article. Technically, Amazon's basic commission is ~15%, and that is in line with other marketplaces.
The issue is that when you pay only that basic referral fee, you won't make a lot of sales. You have to pay more to actually make those sales, which amounts to ~46%. Advertising, fulfillment, and other fees. You have to pay for all of this while being required to keep your price "fair," so you have to raise prices on other platforms to stay afloat.
You could just ignore selling on Amazon altogether, but in the US that's rather unrealistic at this point, and they know that.
The issue is that those policies drive prices up globally, and the article touches on that exact point. Sellers have to absorb Amazon's fees while being required to keep the same price. As a result, their sales outside of Amazon have to factor in those expenses, which drives up prices for those products in general. Even if someone could sell that product at a lower price outside of Amazon, they won't lowball themselves and would rather improve their own margins.
Then, in the EU, you have to factor in fluctuating currency exchange rates and sales tax/VAT set individually by each country. This is another issue altogether.
This specific policy mostly benefits Amazon and its shareholders.
If you think this exists for your benefit, please feel free to bask in the illusion of your wild runaway success.
You would have a point if not for Amazon's exorbitant fees which it expects to be absorbed by the seller and then requires that seller to set the same prices everywhere.
This is clearly manipulative and is regulated outside the US.
It's absolutely insane that this is legal in the US.
This. Simple tagging and/or a dedicated AI news section would be very welcome.
But they actually took some of the risk trying to develop models for Apple Intelligence, and that attempt ended up as a meme. Do you really believe that was a strategic play?
Apple failed in their efforts to build their own models, but they got lucky.
China is doing all the R&D for free and the whole thing turned out to be unprofitable anyway.
For older hardware: Void Linux, Xubuntu or maybe Linux Mint Xfce. If you want something up to date that needs to be online.
AntiX and Puppy Linux are a bit too rough, in my opinion. I'd rather leave the machine with some fully updated old Windows version designed for that hardware, offline. Works very well for retro gaming, ripping CDs and stuff like that.
Chrome is the new Internet Exploder.
Use anything with built-in adblock-rust.
Ironically, their Ad/Business tools are the worst offenders.
I need to use them but I feel like I'm in a perpetual state of fighting their software.
And the valuation is completely justified in this case.
Posted by
mgrunwald_ on Wednesday June 03, 2026 @10:58AM
from the theverge.com dept.
I've seen too many of those privacy-respecting alternative social networks. Problem is that you'll post things only for yourself, which may be cool if you don't care about actually having friends there.
I don't think it's only marketing. OpenAI had the advantage of being first to the market, and in the beginning of the race it seemed that the future belongs to them. Then came the bad PR and unpredictable quality of their main product.
For general use, ChatGPT's answers have gotten worse over the last year. I abandoned it.
I ran this multiple times through GPT-4 and every single time it arrived at the same conclusion. The data was readily available and pretty clear. GPT-5 insisted that the objectively inferior option was better until I gave it my own benchmark data and it was like "Oh okay nevermind".
Gemini's answer was very opinionated and factually correct, whereas Claude gave a more nuanced answer, which was also very good.
As an example, 2026 GPT doesn't even agree with its 2025 self. Last year I asked it to make a hardware comparison and it correctly identified the objectively better option. Recently I asked again and this time and it got everything completely backwards.
I think it depends on how you define god. Personally, I believe in pantheism. It's a philosophical stance which equates god to the universe and nature itself. It's hard not to believe in something that actually exists.
The Splinter Cell and Deus Ex series for me. The stealth genre has this unique quality that you can approach the gameplay in different ways - more brute force, more stealth, anything in between - and have different outcomes/endings.
I don't think they are happy about burning through money, especially if they can just build their data center somewhere else and avoid the 200% tax. The moment of reckoning is happening in real time, all paid plans for AI models are heavily subsidised and heavy users are absolutely ruining their finances. Do you think they would want to build their data center specifically in a 200% data center tax area to make their financial situation even worse? I don't think so.
Please let's not fool ourselves that AI businesses don't care about costs because they have a lot of money. They burn a lot of money, yes, but they are looking for profitability and they won't achieve it with a 200% tax on top of large energy bills and hardware expenses.
This method works better in a free market. Instead of outright banning things, you simply build a system that encourages/disencourages specific things and it basically runs on autopilot.
Flatpak project maintainers, please do not do that. Leave Flatpak universally accessible. I like my alternative Linux distros without systemd.
Exactly. I remember the beginnings of ChatGPT, OpenAI looked like the future and Google had Bard, which was not very good. It looked like Google was soon to become irrelevant. Fast forward to today and they have a lot of great products in this space powered by their own custom AI chips.
Thanks for the comment, I'm glad you like it!
Yes, the pre-release is intended for testing purposes, so thanks for bringing the battery health issue to my attention. It is calculated from the the battery's reported design capacity and current capacity, but the reported values seem to be unreliable across different systems.
The password generator suggestion is interesting, but I intentionally gave the user only one password generator option in the base version of the app - the most secure one :)
I'm glad you like it!
I might consider adding this in a future release, although at this point I'm focusing on keeping the app's footprint minimal, so every feature addition has to be carefully considered :)
I understand your skepticism, but as the original poster and creator of this I am actually a little surprised, too!