So backups are like insurance, it's all about risk, you're right. If a user deletes a record, then a local copy is a great way to restore it quickly. If the computer turns off, you can't do anything. Also if you make a copy once a day and someone created and destroyed a record in that day, you can't restore it. So in practice you use multiple layers, like hourly local copy, daily off the computer, weekly out of the building, or whatever but none of that is ever guaranteed to work.

That all said, if you plan for the 99% and the 1% happens, that's still part of your plan. But if you're in a car wreck, and your auto insurance won't pay out for car wrecks... we say you didn't have insurance, because that's not a socially acceptable level of risk. Similarly, local backups on one computer, great but if that's it and the computer breaks, uhhhh, you were obviously underinsured and should have known better, because it's not a socially accepted risk posture.

The fuck I can't, it's like a dude napping behind the wheel in his Tesla. I'm blaming the shit out of the driver, the car, and the people that made it. There's too much negligence all around to ignore any of it.

Do you understand to get to this point you have to have a bearer token with delete the primary database and data privileges, sitting on someone's laptop, with direct access to the infrastructure, which is on the Internet.. so everywhere.

Then you have to have a SaaS vendor who's service will do that destructive act immediately, instead of turning it off and starting a 24h timer.

Now we're very close to blaming the computer.

I'm not one to victim blame, but that _whole_ situation is "as intended", and way waaaaaaaaay too common. We all need to do better, and yes AI tool vendors have a big part to play in fixing this too. That XKCD meme with the giant stack of devops dependencies and a little js library holding it all up, it feels like some moron with a chatbot is going to blow it over and it's going to be way to fucking easy to say don't blame the user... the computer did that.. the vendor did it.. etc. Fuck that I'm not having it, I'm blaming every last person that touched any part of that stupidity, and that's basically all of us, I'll take it.

Mmm nah. Though if he's smoking his own supply, yes. If he's spinning this to deflect blame from his company, then as gross at it appears, he's doing his job well. If there's no repeat anyway.

It fits a pattern. Firing and harassing military lawyers because they want to replace them with yes-men. It doesn't get any more plain than this. When your boss fires the company lawyer, not for past performance, but ahead of time because he might not like what's coming, what the fuck is happening.

https://apnews.com/article/pen...

https://www.govexec.com/defens...

Wait... in rural BFE kids might start dirt biking at like 6, 7. We didn't have speedometers but ... 50cc is just the STARTING point, and those have no problem going 20. And off road, where at any speed you learn scanning the ground in front of you like a hawk.

E-bikes come in every size and are available in places dirt bikes aren't. There's just no reason to assume kids are inexperienced with speed like that. I mean maybe.. everyone starts somewhere. The way you're looking at it as a committed bicyclist, like you exercise or commute or mountain bike downhill, more grown up stuff and experience that young people mostly don't. I can tell you there's a path for very young people through motorsports. Since these lines are blurring, I wouldn't make assumptions. Maybe some 14 yr old just got a Surron last summer and is a total dbag with it, but he could be like a suburban version of the country kid that started on a 50cc trail bike is all I'm saying, might have spent half his life riding already, and a 125 is getting small.. there's certainly going to be more of that because it's more accessible

How is THAT the problem. The specifications give manufacturers and regulators a baseline.

Just take a plain old dirt bike for example, needs a CA legal spark arresting muffler. Any meathead with a screwdriver can remove it or replace a muffler. The important thing is most trail bikes now have spark arrestors and if you fuck with it and THEN cause a fire, there's a different liability picture. You didn't just cause the problem you did it with a level of negligence.

Unless there is actually a pervasive problem with people boosting their 20hp bike to 30hp or what t f ever, and riding on 20hp restricted trails, and THEN truly being idiots with it, then it's not remotely an enforcement problem. First you need some classification, just to do that first part, to even make a trail for some kinds of mixed use, and others for go-fast. Yah anyone can break the rules, no shit... just don't get caught and don't make it a problem.

If you were driving a Surron down the sidewalk to the office everyday, yes please do. And stay off the sidewalk, if it needed to be said, idiot.

Obviously they'd rather be pulling easy overtime parked in front of some construction job than counting fish, but you still need game wardens. We had to register our dirt bikes out in rural BFE and they'd catch us if we didn't, or sit at the empty truck parked next to the gravel pit and wait you out.

Ignoring the problems doesn't help. Registration and classification need to happen, and it will happen in every state, it's a matter of time. E-bikes are in a strange spot, not likely to start a grass fire so some old rules and conventions shouldn't apply, but they can wreck a trail as bad as anything else, so some rules should. They're not obnoxious sounding so that should open some unconventional paths, but they also have power output literally anywhere between pedal assist and motocross while sort of looking and sounding the same, and weighing anywhere in between. So in terms of damage they could do to trails, lawns, pedestrians, other trail users etc, it's all over the place.

I get that we shouldn't race to rules and regulations right away, because it might have stymied all that variety that we have right now, bikes would be made for different CA legal buckets and that could kill a lot of the different options we have. But dude.. putting a plate on them is the most obvious first step, it's so the cops DON'T have to chase down every dumbass kid, we can take a picture of them cutting holes in the soccer field and get the ball rolling on trespassing charges quicker. Go look at any Facebook group for suburban anywhere and you'll see what the problem is.

At the same time, we need more e-bike friendly trail options out in suburban areas, because it's a great way to get kids outside and active, and off the places we don't want them using these things. Like what we do for skaters.

So this isn't a criminal trial, and the bar to find the defendant liable is much lower than it would be in a criminal trial right? More likely than not, as opposed to beyond reasonable doubt. 51% sure, instead of 90 something right. In other words, easier on Musk than it would be on a federal prosecutor.

The question was how many people have to hate a public figure for them to effectively lose their due process rights, when the defendant requests a jury (because a well hated plaintiff can opt not to request one)? So how strong or how much hate does there need to be for ... eight to twelve people to unanimously ignore objective reality, when evaluating legal claims broken down to them by a judge? And for the judge to be unable to manage this?

I mean it's a fun question. Pick some extremes and test it. Trump, the plaintiff, sues Taylor Swift for uhh.. defamation? In... New York City.
How strong does then TDS need to be for a jury to unanimously flat out ignore reality when asked things like "did the defendant communicate the statement to at least one third party?", "Was it presented as a factual assertion", "is the statement false?", did the defendant know it was false, etc. You don't think a group of people can get together even with those extremes and decide yes and no fairly after hearing both sides?

Look, like it or not, Musk will have a fair trial. I think it takes a whole lot more than what you're insinuating for a plaintiff that has actually been wronged in some way to not be heard. All the examples I can come up with where it might be hard are just so bad they're almost funny.. like could OJ truly get an impartial jury trial to decide he should get his "stolen" swag back? I hate to say it but who gives a fuck. And it's not even about the plaintiff because psychologically, it comes down to did the defendant do the dirty or not because people are really unlikely to overlook you know.. brazen daylight robbery, or financial crime, or something immoral just because they want to stick it to the plaintiff. But Christ if they did, it's a fair question man, how bad do you have to be, but also how frivolous sounding their claims are.

Well ok, sure, but the Vision Pro is no iPhone 3. They'd have to make a Vision Air, a Vision Pro 2, a Vision 3 Neo, or whatever for this thing to pay off in the end most likely. Iterate and get the cost down, find and support 3rd party developers that do things they never even intended. Not saying they will, but this is how it went down with the iPhone ain't it, it ended up a ways from where it started. Then there's the Apple TV that kind of sat there quietly and never really changed, and after a very loooong time ended up sort of complementing a large new part of their business. So if someone told me they completely abandoned this product line, the whole idea, I'd believe it, but I'd also believe if they had some convoluted ten year plan, that they have the resources to stick with it and end up ... somewhere they probably didn't intend but they have the resources to get there.

Something to consider, by public good, OpenAI's original charter is to develop AGI that benefits all of humanity, not just a small group. Not to share an open source free to use coding model.

Small problem, as noble as their non-profit charter sounds, I don't have the faintest idea what that actually looks like. What I do know, is providing a free coding model is not it, and neither is selling lots of coding model subscriptions. One of those could at least possibly fund the actual mission, one is sort of pointless. I don't see a conflict with either.

Either we say this won't lead to AGI and spin it off to raise money for the mission, or shut it down and give away the work and say mission accomplished? Or stick around with their hand out for donations to run huge data centers for the sake of providing cheap AI for those in need? See we can't say that with a straight face. No matter how I look at it, I can't see selling coding model subscriptions, even if we don't like the idea of, I can't see it in conflict with the development of AGI that benefits all of humanity. Whatever that is, it's not through the door of LLM coding models, even though the research is related.

I'm not a lawyer, but I think part of the deal for non-profit status, because it's tax advantaged, is that they do need to have a specific purpose within some guidelines. That's probably what your down mods are from, but I think we agree that maximizing the value of the held intellectual property is NOT required of a non-profit, and it's fully within a non-profit's rights to license or sell their property, if that is supported by what they do actually have to do.

Hah, I knew dual licensing would perk some ears. But, if it is the objective, to earmark a donation for a specific purpose, there isn't a vehicle like copyright to do that with. That I know of anyway, but say there was, it can go two ways. For example, there's donating to the Girl Scouts explicitly to further the Girl Scouts actual, legally required or not, moral purpose for being... and then there's making a donation to the Girl Scouts that's somehow earmarked to expand cookie production. Which is definitely not in any way the purpose of that organization. Which did Musk want, it sounds like he's butthurt he was cut out of a stake in the cookie production spinoff and thinks nobody should have it then.

If we look at something like MySQL, I can at least see a conflict of interest between the free project and the commercial product sides. If we divest the cookie business from the Girl Scouts, I don't see a conflict of interest there between Girl Scouts and a cookie company, provided they find a new way to teach entrepreneurialism. I also don't see a conflict of interest between OpenAI's original non-profit mission to develop AGI that benefits all humanity, and selling as many codex subs to maximize profit from commercial vibe coding. The two objectives don't conflict, putting aside what we think AGI is, we know it's not that, and if that raises money and furthers actual AGI development and the non-profit's legal, moral, ethical etc requirements, than there's nothing wrong spinning off the IP in my eyes.

OpenAI Foundation, the non-profit. It wasn't sold, isn't being sold, and nobody is trying to sell it. The "valuation" Musk is referring to must be the $$$ value that OpenAI has, that definitely can be sold. When we say something doesn't make sense, because it can't happen, and something real was misidentified, can we come back to the real thing? For everyone's benefit I'll assume Mr Musk was confused.

No need to play the "if you could buy" game with the Salvation Army, let's use the Girl Scouts.

The non-profit, Girl Scouts of the USA, their mission is NOT to grow a cookie production empire.
Its Girl Scouts Cookie Program is a business.
The Girl Scouts cookie business can be valued like any other business.
If you can sell a cookie, you can sell a cookie business.
And none of this is weird.

Please don't take this as criticism for how the Girl Scouts operates, but retaining the cookie program inside the non-profit is not actually the mission. If teaching entrepreneurialism to young people is part of the non-profit mission, they theoretically hold bake sales. The cookie business could at any point be sold or licensed, free to maximize its profits or whatever objective Girl Scouts USA or its new owners have, and its sale or licensing agreement would fund Girl Scouts USA's actual mission. There's nothing weird in any of that, or the Salvation Army licensing its brand to a for-profit company that sells camo shorts, hypothetically. There are a lot of good and honorable reasons to do these things, I'm not saying they should.

If a complaint is that a donation to Girl Scouts USA didn't go to expanding their cookie product lines, THAT is weird. That is what I see here with OpenAI. Someone that avoids talking about the non-profit's actual mission, because that's inconvenient to their argument, and sour they weren't part of the business potential that was spun off. Their mission is to develop AGI that benefits all of humanity. You can call that entire concept BS if you want, it is totally irrelevant because what the mission is NOT is to sell as many codex subscriptions as possible. How can anyone see what's happening as anything other than if can't have it nobody can?

It has revenue, it has value. I'm not following Musk's logic. Say it was a cancer research charity. You make a donation. That money goes fully to the non-profit's mission of cancer research and affordable therapy. They later sell or license some revenue generating IP they created to a controlled subsidiary. That licensing funds the non-profit's governance and its mission, the for profit arm can grow faster from outside investments and grant equity to highly poachable employees. Is the non-profit still following its mission and spending every dollar of donations and revenue on the mission? Yes. Does it need to retain all revenue growth? Why? ... it's a non-profit ...

It's just weird to look at it like you got shafted because your donation didn't "buy" all the maximum earnings potential (that as a non-profit is not even the objective, RIGHT?) and that you wouldn't hold a stake in anyway. This is the same weird argument people make with open source dual license projects, like you give something away to be used for whatever, but someone else does something with it and suddenly they "took" something from you. Nah your stuff is right there doing its thing, it didn't go anywhere, it's still doing exactly what it claimed to do. If you wanted a share of the future earnings you're in the wrong place sorry.

That was later, the iPhone launched with built-in apps and no SDK, no App Store. The answer at the time was web apps. Which, considering the gigantic leap from Opera on a Blackberry to Safari on an iPhone actually wasn't a bad plan.