You're right in spirit, but where is the "power cord" for an AI running in the cloud? Even taking an entire data center offline often isn't adequate when the workloads are explicitly designed to be location-agnostic. We always need to be cautious about giving governments more power, but in this case there's a clear legitimate need to force AI companies to come up with some means by which their AIs can be immediately stopped in their tracks.

The bigger issue here is a kill switch may not be possible in any useful sense. An AI is little more than a description of its topology and a set of weights. Any halfway decent programmer could roll their own CPU-based implementation of even top frontier models, performance aside. And sure, the throughput might be measured in minutes per token, but so what? The answer to "how do I hack $TLA" is likely to be just as useful in five days as it would be in five minutes - Heck, the DOD and DOJ didn't even notice they'd been compromised for months, and they're supposed to be some of the most secure organizations in the modern world.

Oh, I totally get that point, and agree! But my point remains: We don't need to replace Teams-as-a-whole with a craptastic monolithic of mismatched functionality shoehorned into one poorly-behaved application. We need a messaging app, halfway decent teleconferencing, and a collaborative document management platform. And I guess some form of locally mountable cloud storage, if you're a fan of giving your personal data to completely untrustworthy third parties.

I think the bigger problem is trying to replace all of Teams in the first place.

Just give us Skype back! Sharepoint already exists and works fine in a browser (sure, even Edge if we want to go there); and I actively go out of my way to remove OneDrive every time MS shoves it back down our throats. Bam, there goes 90% of what's wrong with Teams.

IANAL, but perhaps someone who is could explain why on Earth they went for antitrust rather than something with a lower bar like tortuous interference?

As I understand it, the former is virtually impossible to win even under the best of conditions; the criteria for the latter is practically Google's own defense in this case - "Yeah, we intentionally destroyed any shot at you ever getting another customer solely to boost our own ad revenue, by deceptively offering you a free carrot while we stole the whole farm out from under you - But we don't have a contract with you, so nyah-nyah!"

Fortunately, societal expectations have changed radically over the past few decades. In the 1950s virtually nobody was voluntarily child-free. In the '70s that rose for the first time to roughly 10%. By 2010 it was 20%. Today (well, in 2023) it's 29% and still appears to be rising rapidly.

It's one thing to have a vague sense of despair for the future of humanity; it's another matter entirely to act on that feeling if it means your family and most friends will disown you. When I chose not to have kids 30ish years ago, fortunately I had fairly supportive parents (though the family elders nag me at every gathering even now); but I still watched most of my childhood friends basically drop off the map as soon as they had kids. Even at work, where reproductive status is theoretically a federally protected class, I would always get the worst shifts and even when I became a 9-to-5'er, the parents constantly check out early to pick their kids up from random events, and they always get first pick of PTO because boo-fuckin'-hoo they have a harder time scheduling things.

By contrast, today a third or more of someone's high-school friends are likely to remain after the breeders pair up and vanish from circulation. Parents have come to realize they may not ever get grandkids, and if they really want some, they'd better pop out a few extras of their own to boost the odds. And work... Still tries to play those same games, but I'm old enough now to put my foot down and demand they send a copy of any discriminatory policies in writing to my lawyer - To be clear, I'm not actually a litigious bastard, but those magic words are enough to make even the most worthless of HR drones stop parroting "RAAAWK! Employee Handbook, comply, comply!" and engage the brain long enough to realize what they're saying is blatantly illegal.

Whether or not the world is worse today than it was near the end of the Cold War, people are finally (somewhat) free to follow their conscience.

Ditto. I was a fairly prolific Redditor for the past decade, and stopped cold-turkey the day they stopped allowing anonymous access to Old.

That said, that was just the final straw. Reddit has been going downhill for a few years now - Right around the same time bots outnumbered the humans, and AI has only sped that up by an order of magnitude. The "Dead Internet" is already here.

Not to mention Reddit's moderation has become so blatantly partisan it's impossible to discuss anything but the most banal of topics. Even mentioning I/P without strongly suggesting we'd really love to fellate Bibi for his greatest hits (and not on schools and hospitals) is all but a guaranteed timeout.

Reddit was a great site, as close to an all-topic old-school forum as we could ask for. "Was".

For someone who considers the MoCA "a very hard test", it's not all that unreasonable to consider an obsequious oracle "super intelligent".

Of course, we could say the same about a slightly brain-damaged labrador retriever, but they aren't setting national and global policy regarding the most likely cause of human extinction over the next few decades.

I kind of agree it's increasingly annoying to find dead space on a titlebar, and I absolutely loathe Microsoft's use of the title bar to stick crap I have never and will never use. I don't search enough to waste half the title bar on a search bar, I couldn't care less who I'm logged in as, and "Saved to this PC" is nothing more than Microsoft flipping us off for having zero interest in automatically saving everything OneDrive.

At the same time, I really do like using the title bar as a tab bar. It's a decent compromise insofar as they're conceptually like mini title bars for each page we have open.

Round controls, however, need to die yesterday. Rounded, sure, but having a full semicircle wasted at each end of a button is just a gratuitous waste of space. "Compact" is fundamentally incompatible with "round".

Am I missing something here? Application Menu -> Settings -> Extensions&Themes -> Themes. You can choose a few common defaults, customize your own, or download thousands of them from addons.mozilla.org.

You're the third comment implying we can no longer apply (or even make) custom themes, yet the linked announcement bluntly states exactly the opposite and implies they're increasing the number of themes available. Admittedly I run ESR rather than the bleeding edge, but we're only talking six months old, not some antique from the late '90s.

You don't need to edit your user prefs to disable dark mode, themes are right there in the main settings.

That said, the vast majority of people using dark mode will consider some purple highlights in dark mode far, far less annoying than switching to light mode.

Personally, I prefer most of my "non-content" screen elements to be low-contrast grayscale and, wherever possible, dark grey window backgrounds. Purple is an annoying violation of that preference, but still beats bright white backgrounds and obnoxious Candy-Crush-like icons / controls / borders by a mile.

So last week's "summit" went that well, eh?

So we're going to use an AI to monitor AIs to make sure they're not doing anything naughty? Pretty sure I've seen that movie already...

Not to mention, who exactly pays for Sentry to run? Golly, if some completely random AI hardware manufacturer were to somehow profit off designing and selling a scheme to waste extra tokens on yet another layer of supervision, that would almost look just a wee bit unethical!

That wasn't quite the response I expected - I guess I'm not entirely clear on whether you were mocking those guardrails, or presenting them as legitimately (albeit perversely) meaningful.

Either way, my central point was "why not both". It's entirely possible to be neutral, in the sense of calculating the mean (in the internal thought-vector sense) of all available viewpoints. If, however, 99% of sources are talking positively about some trashy conspiracy theory - Any version of "consensus" is itself going to lean pretty far into wingnut territory. In that regard, I think we might actually be agreeing?

We can meaningfully ask the question "For each location a robotaxi has been involved in some sort of incident, how does the incident rate for both humans and other robotaxies comapre?"

That's not a meaningless number, and it's entirely possible Waymo really is doing better than humans - But only in situations Waymo gets to cherry-pick simply by enforcing their designated service areas. Meanwhile, let me know when a Waymo can successfully navigate overgrown single-track dirt roads common in my area. Anyone know offhand if they have moose in the Philippines?

I'm a fan of drastically slowing AI development until we've solved the alignment problem, but a kill switch ain't it.

What would a kill switch really do? An AI is fundamentally nothing more than a description of it's topology and a set of a few billion weights. Implementing that topology and running those weights is the easy part and any coder worth a damn could roll their own (performance aside) with minimal effort - without bothering with a kill switch, if they so choose.

The only way to even come close to enforcing a kill order would be for every sufficiently-performant processing platform to enforce some kind of blacklist of known-evil weights in a way that's resilient across different quantization levels. And even then, simply running pre-killswitch drivers would still be a trivial way to circumvent any such restrictions entirely. Or for the truly determined, you can run even top frontier models on vastly underpowered hardware like a mid-range desktop CPU. Sure, your throughput might be measured in minutes per token, but realtime feedback isn't going to change the answer to "how do I hack the Gibson". An accurate answer a few days from now is likely to be just as useful in the vast majority of situations.