I would expect that your whole code base would be then used as training data.

I think that it is only a matter of time before they will start reselling these data as exfiltrated IP to whoever will be interested.

One day there will be a massive data leak and these backdoors are going to be closed. Not because privacy of some Joe Smith or exposing corruption, but because everyone will be able to read raunchy DMs of random politicians with their mistresses.

It will be funny when somebody will figure out that contractors watching Chinese nationals are actually Chinese spies.

Yet another argument for running your own local models. And if they are scanning, they are definitely using all your data for training or outright IP theft, no matter what they claim. Current approach is just vacuum cleaner for your data.

I bet on natural gas powerplants will start being built together with data centers.

Ironically xAI is already doing this.

Permanent until Trump will talk to somebody else than Navarro about these tariffs and will be promptly lifted.

The problem with current setup is attempt to strongarm tariffs onto seller, who will then promptly either stop sending any packages (because that's not how the system works) or bill the tariff back to customer explicitly - Here is your invoice for shipment and here is your invoice for tariff. You are just creating more bureaucracy at this point for no benefit.

Just compile .NET WPF in Avalonia and be done with it.

> It’s really wild the degree to which Musk’s Nazi salute has ricocheted around Europe

This is because Americans does not have places in their country with mementos like "In this building Gestapo tortured 20 people to death ..." or "659 allied soldiers died liberating this city" or "In this complex of buildings over 500 thousand people were gassed to death"

For Musk it was a funny free speech thing. For Europeans it was very personal and very painful part of history.

English is actually a weird language without genders in nouns. I.e. in Slavic languages you can say "male software engineer" with word "vyvojar" and "female software engineer" with a word "vyvojarka" and then lot of grammar is built atop of this fact.

Job listing are then trying to use something like "vyvojar/ka" to signify that both genders are sought for, but there is nothing like that in English, so you will get translation as "software engineer (all genders)" instead of using just "software engineer"

fe/male perhaps

And this is why local run models are absolutely necessary. Sure Claude is better than whatever you can run locally, but to avoid being eavesdropped on every keystroke, just buy older enterprise server with enough compute for 3k USD and run similar model there.

Perhaps design with public model and then convert to a local one.