Great article, thanks for sharing it OP.

For example, the Anthropic researchers who located the concept of the Golden Gate Bridge within Claude didn’t just identify the regions of the model that lit up when the bridge was on Claude’s mind. They took a profound next step: They tweaked the model so that the weights in those regions were 10 times stronger than they’d been before. This form of “clamping” the model weights meant that even if the Golden Gate Bridge was not mentioned in a given prompt, or was not somehow a natural answer to a user’s question on the basis of its regular training and tuning, the activations of those regions would always be high.

The result? Clamping those weights enough made Claude obsess about the Golden Gate Bridge. As Anthropic described it:

If you ask this “Golden Gate Claude” how to spend $10, it will recommend using it to drive across the Golden Gate Bridge and pay the toll. If you ask it to write a love story, it’ll tell you a tale of a car who can’t wait to cross its beloved bridge on a foggy day. If you ask it what it imagines it looks like, it will likely tell you that it imagines it looks like the Golden Gate Bridge.

Okay, now imagine you’re Elon Musk and you really want to change hearts and minds on the topic of, for example, white supremacy. AI chatbots have the potential to fundamentally change how a wide swath of people perceive reality.

If we think the reality distortion bubble is bad now (MAGAsphere, etc), how bad will things get when people implicitly trust the output from these models and the underlying process by which the model decides how to present information is weighted towards particular ideologies? Considering the rest of the article, which explores the way in which chatbots attempt to create a profile for the user and serve different content based on that profile, now it will be even easier to identify those most susceptible to mis/disinformation and deliver it with a cheery tone.

How might we, as a society, create a process for conducting oversight for these “tools”? We need a cohesive approach that can be explained to policymakers in a way that will call them to action on this issue.

on device

scam detection

I know I’ll be downvoted into oblivion as I can hardly believe I’ve formed this opinion myself, but tbh this is a good application for some of this AI tech.

Anecdotally, a friend of mine grew up well-off; from an immigrant family but their parents were educated and in a lucrative profession so he always went to private schools etc. Fast forward to about 10 years after all the kids moved out; the parents had divorced amicably and his mom had a sizeable retirement along with the payout she had from the divorce. In the 7 figures - she never had to worry about money.

Anywho, mom ran into some medical issues so the kids had to get involved with her finances again, as she couldn’t do it herself. Turns out that over the course of months or years, mom had been getting scammed to the tune of tens of thousands of dollars at a time, to the point where she had actually taken out a mortgage on the home she previously owned outright. They’re still sorting things out but the number he has tossed out in the past is ~$1.4M that got wired overseas and is just… gone now.

So yes, I probably won’t turn this feature on myself, but for the tens of millions of uneducated and inept people out there, this could genuinely make a difference in avoiding some catastrophic outcomes. It certainly isn’t a perfect solution, but I suspect my friend would rate it as much better than nothing, and I would argue that this falls short of being “strictly evil”.

I was in a similar boat and have found modern Linux to be somewhere between Windows XP and Windows 10 in terms of convenience and having it “just work”. However, I reckon I’ve spent less time troubleshooting than I would spend raging at the bullshit Microsoft keeps trying to shove down your throat in Windows. On balance I’m counting it as a win, and I suspect you will too.

Also don’t forget they’re planning an IPO this year to massively accelerate the enshittification. Surely their investors wouldn’t demand that this biometric data be monetized.

Update: it’s a great game. I’m not even a huge hockey fan, but it sucked me right in.

Tape to tape? Oh that’s new to me… Thanks friend

When you see people on Lemmy casually joking about eating the rich, they’re talking about you.

The term op-ed is literally short for “opposite the editorial page”, and the idea is that it is an opportunity for opposing viewpoints to be brought into the discussion for consideration. The fact that the author is a moron does not besmirch the quality of the Guardian writ large.

Motherfucker how long do you think humans have been around for? Okay, sure, soap has been around for like 6k years, wow, such a long time. Modern humans have been fumbling around figuring shit out like an untrained ai model for the last 60k-160k years. Quick math shows us that for somewhere between 54k and 154k years of our history as a species, no soap. MILLENIA!

Drinking water: you’re making a distinction between clean, as in unpolluted by chemicals and other substances, and cleaned drinking water, which has been processed by humans to make it fit to drink with a lower risk of causing illness. Clean water has obviously been around for eons, but cleaned water, as I believe OP was describing, is a much more modern concept.

Showers: “try skipping a shower for a day”? Motherfucking neckbeard no, shower every fucking day. Try it, people might find you somewhat less repulsive until you open your mouth.

Agreed. As with most things, the answer is neither black, nor white, but some shade of grey.

I disagree. Most people are able to self-regulate in more instances than you give credit for. In instances where someone is unable to do so, I would agree that a full prohibition probably makes sense; such as with alcoholics who are unable to stop themselves, but this is not the case for most people.

Interruptions - fear not the notification, for you hold the power to calibrate your notifications to the level that suits you. I’m particularly aggressive about disabling notifications and especially notification categories that do not add value, and you can too.

Resisting - many humans have reported experiencing even more enjoyment engaging in activities they enjoy if they delay their engagement until meeting a predetermined goal. If you know that scrolling through your favorite app for 5 min will give you a dopamine hit, you can choose to delay using the app until after you’ve completed a chore, for example, and that may offer enough incentive to complete more tasks than you might otherwise be able to perform.

I’ve been assuming it’s because they truly have no idea how this tech works

Yeah it’s already a problem, but if we start ramming satellites against each other I could imagine the problem getting to a point where we have trouble launching things into space. The real issue is the relative velocity that orbiting objects have. If you’re curious, the Wikipedia article has a really cool infographic and the section on hazards is pretty interesting if you like space stuff.

At some point we will have enough shit orbiting the planet that we won’t be able to launch anything into space for decades. That will be a good time.

I need a gif where Scooby Doo removes the Librewolf logo and there’s a Firefox logo underneath.

You must recognize that there is no Librewolf without Firefox, right? In fact, Librewolf even says in their privacy policy that you should also refer to the Firefox Privacy Policy because they can’t be certain that their browser won’t ever try to send data to Mozilla.

I’m not saying this to deter you from using Librewolf. If it works for you then that’s awesome. It just made me chuckle when you said that you ended your friendship with Firefox and ran into the warm embrace of… Firefox with different default settings.

In any case, all I’m trying to communicate is that Firefox and all of its many forks are fundamentally reliant on Mozilla and its ability to continue updating Firefox. That means Mozilla needs a sustainable business model, and that we can’t all simply abandon our relationship with Mozilla for a tool that is dependent on the work that Mozilla does.

Donating has always been an option

Omg same. There are literally dozens of us.

Ooh, new Creeper World? Underrated games, though the concept was tough for me to get my head wrapped around for a little bit.

Feels like deciding in 2010 between Twitter and Reddit in some ways…

Steam has voice channels? 🤯