That’s great advice, when you don’t have to explain to your elderly parents why you need to migrate their email for the second time in six months.

The only reason I haven’t cancelled yet is because I moved my family, including my parents, onto a family plan with them, and that’s going to be a lot of work to move everyone again.

But this shit fucking sucks. I recommended this product to my loved ones because it’s supposed to be trustworthy and secure, and now I find out it’s neither of those things.

If you’re a Proton customer right now and not in a position to simply cancel your account, submit a support ticket laying out your concerns. I got a detailed and lengthy response back within a day (one that was clearly not a form letter either, it directly addressed all the points I raised). Flooding their support channels will make it clear that they are in some serious shit here.

Write to the Proton Foundation too. Demand that Yen be removed from the board and the company, and that serious changes be made to the leadership at Proton to ensure this never happens again.

Sure, those are all things the government could do in a world where this wasn’t the intended outcome.

NASA is being fed to SpaceX like a prime rib.

Meta does have some very specific issues with their corporate governance in that regard. Mark’s contact basically makes him impossible to fire as CEO. It’s an unusual situation that should have set off alarm bells for investors years ago.

Now that people are shunting easier requests to local models and sending the harder problems out, the hyper scalers have less income, so they have to charge more.

While you’re 100% correct about the fact that this is happening, and it is a serious problem for the hyper-scalers, it’s really, really important to note that even before this started happening, LLMs were already hilariously unprofitable, both on the software and the hardware side. For the AI companies, there is simply no way to charge people what the models actually cost to run and still turn a profit. On the data-centre side, the turnover on burned out GPUs is going to annihilate any hope of profitability. They’re treating as capex what is essentially a consumable and their business models simply do not account for this reality.

The behaviour you’re identifying here is basically just a swift kick in the guts to someone who was already bleeding to death from several bullet wounds. Seriously unhelpful, but not remotely the impending cause of death.

Basically, yes.

This comes back to the basic problem with LLMs as a business proposition; any digital service can either be cheap and frivolous, or expensive and necessary. You can even, if you do things right, be cheap and necessary. But you cannot ever be expensive and frivolous.

LLMs were supposed to be expensive and necessary, but it turns out they’re just so unreliable that they simply cannot perform an essential role. They must always be disposable; fun, or useful, but never essential.

Which leaves only cheap and frivolous as an option. And they’re just not cheap, nor does it seem likely that they ever will be.

So you get Character.ai filling chapter 11, because people simply will not ever pay what it actually costs to run these models just so that they can have a computer program pretend to be Lois Griffin for them.

My honest recommendation here would be to use Notesnook instead. I used Joplin for a while, and when I found Notesnook I switched immediately. It’s just flat out better at basically everything Joplin does, and it’s still open source. No need to run syncthing, use the notesnook server to sync and backup everything.

I’m a huge fan of Caddy and I wish more people would try it. The utter simplicity of the config file is breathtaking when you compare it with Apache or Nginx. Stuff that takes twenty or thirty lines in other webservers becomes just one in Caddy.