modulus

Interesting article, and I definitely agree I prefer clear instructions when those are possible.

I only have an objection. When it’s said that no matter how well chatbots behave, it’s bad design, and that they’re being used to substitute expensive people; well, expensive people’s interface is chatting too. So in that regard I’m not sure there’s a meaningful difference. Obviously there is if the chatbot is badly behaved, but the article says that it’s a problem even setting that aside.

It’s an NVDA add-on. It substitutes role announcements with specific 3d located sounds. And fortunately it does work. I should change the manifest to indicate it.

I’m going to be trying that today. I hope my Unspoken version still runs on this.

Wow, it’s like he chose those examples on purpose to make his argument as ridiculous as possible: open borders (except for all the people forbidden to leave), regular elections (except now they’re indefinitely postponed)…

Get your DeepSeek3 and r1 weights before it’s illegal!

One of the things you’re missing is the same techniques are applicable to multimodality. They’ve already released a multimodal model: https://seekingalpha.com/news/4398945-deepseek-releases-open-source-ai-multimodal-model-janus-pro-7b

I see some people are having issues with the scenario, but it’s not as impossible as it seems. The key is that Newtonian mechanics are in principle time-reversible. If a system got to a state one way, it can get back to the state it was by running it backwards, so to speak. A ball going down an inclined plain with a given kinetic energy could be going up that inclined plain up to the top with that same amount of energy.

The problem with these systems is, it’s possible to impel the right amount of force on a mobile so that it goes through a path and then stops. But since there is time reversibility, it should be possible for the mobile to spontaneously start moving from that stopping point and draw the same path.

Other weird similar cases are the so-called space invader (particle going to infinity, and therefore spontaneously appearing in reverse) and some strange n-body problem cases.

I kept giving Mozilla the benefit of the doubt and telling myself things weren’t so bad.

I was wrong.

I’ll continue using Firefox because it’s the least bad option, but I can’t advocate for it in good faith anymore, and I don’t expect it to last long with this orientation.

So it goes.

So it would still help optimising persuasion at scale (also known as lying to people to best et them to act against their interest). Why is this a good thing again?

what do I think the history is? A record of the sites I visited.

What do I think the history isn’t? A correlated record of which advertisements I’ve been exposed to, and which conversions I’ve made, that gets sent to people who are not me.

Pretty relevant distinction. One thing is me tracking myself, another thing is this tracking being sent to others, no matter how purportedly trustworthy.

I’d like people to STOP PRETENDING that the only plausible reason why someone doesn’t agree with this is that we don’t understand it. Yes, I understand what this does. The browser tracks which advertisements have been visited, the advertiser indicates to the browser when a conversion action happens, and the browser sends this information to a third-party aggregator which uses differential techniques to make it infeasible to deanonymise specific users. Do I get a pass?

Yes, this is actively collaborating with advertising. It is, in the words of Mozilla, useful to advertisers. It involves going down a level from being tracked by remote sites to being tracked by my own browser, running on my own machine. Setting aside the issues of institutional design and the possibility for data leaks, it’s still helping people whose business is to convince me to do things against my interest, to do so more effectively.

I don’t blame Mozilla for not single-handedly ending advertising online. That’s too much to expect from anyone. But they could at least avoid active collaboration with the enterprise. And if they’re going to engage in it, they should at the very least warn their users.

Whatever opinion you may have of advertising as an economic model, it’s a powerful industry that’s not going to pack up and go away.

Fuck that. Not if we don’t make it. That’s precisely the point. Do not comply. Do not submit. Never. Advertising is contrary to the interests of humanity. You’re never going to convince me becoming a collaborator for a hypothetically less pernicious form is the right course of action. Never. No quarter.

We’ve been collaborating with Meta on this,

That makes it even worse.

any successful mechanism will need to be actually useful to advertisers,

And therefore inimical to humanity in general and users in particular.

Digital advertising is not going away,

Not with that attitude.

but the surveillance parts could actually go away

Aggregate surveillance is still surveillance. It is still intrusive, it still leverages aggregate human behaviour in order to harm humans by convincing them to do things against their own interest and in the interest of the advertiser.

This is supposedly an experiment. You’ve decided to run an experiment on users without consent. And you still think this is the right thing–since you claim the default is the correct behaviour.

I cannot trust this.

It’s hard when I don’t get told about it and find by chance.

Yes, for example I donate to thunderbird since I find it useful. And I wouldn’t mind donating to Firefox either provided they wouldn’t do this sort of fuckery.

though in the long run we need to overturn capitalism of course, and that an economic model is viable doesn’t mean we should sustain it or justify it.

It depends, but mostly no. And if that means some sites are not economically possible, so be it.

This is bullshit. The total amount of advertising I want is zero. The total amount I want of tracking is zero. The total amount of experiments I want run on my data without consent is, guess, zero.