paddw

imo, if you don't want the government to track you, the battle is already lost, but there is probably room to limit what private companies are allowed to do on the regulatory front.

Isn't that the point?

The article does not answer the question. (I don't believe there is a straightforward best answer but the article does not so much as make a conjecture beyond the most obvious prerequisites.)

This isn't "the military" doing anything, these are individuals taking their own initiative. (Military probably doesn't mind though.)

Are droves of young men really being driven into the ranks of the army in frenzies of desire? I somehow doubt it.

I think what makes this a "fuck you" pattern is the fact that you are given a "teaser" of the content before being forced to log in.

Like, if IG forced you to login to see anything, you might dislike that, but it wouldn't feel manipulative.

> Saying no to a senior leader at work is almost unthinkable, even laughable, for many people. However, when saying yes is going to compromise your ability to make the highest level of contribution to your work, it is also your obligation.

What kind of world do the people who write this stuff live in?

Company exists to return capital to shareholders. In the absence of a dividend, that has to be from stock price going up. If neither happens, leadership will be changed. Efficient market hypothesis is also nonsense.

The most recent flop was new IP

> I do judge that dad who convinced his teenage son to risk his life. That was absolutely evil.

Moreso, he was likely foolish in considering the actual extent of the risk, like everyone on the sub.

https://wasi.dev/ allegedly. Development seems slow though.

There's no way Discord is profitable.

"You hate these guys right? How about you make them give us some cash??"

You don't think there's any political reason they are going after Amazon rather than, say, the New York Times? Just the impartial watchmen of Democracy studiously promoting the adherence to regulation here?

I think OpenAI's moat is a large corpus of well-cleaned training data. There is no brilliant insight behind the model architecture for GPT-4, but rather 100s of small incremental improvements upon previous model iterations. Eventually, large companies will be able to catch up. Open source should also be able to, in theory, all though it will be hard for things to come together.

There are so many of these processes which are so difficult that they border on the outright criminal, but the FTC goes after Amazon because it is a big political target to hit. Maybe from a utilitarian perspective it makes sense.

The $140,000 constitutes back-wages and damages. I agree it is still too low.

Well, the problem is you can easily trade gold for cash because it is socially acceptable to do so. If Bitcoin became outlawed, it would be very hard to exchange Bitcoin for cash, since it has no extrinsic value outside of being a financial asset (as compared to, for example, barrels of crude oil). That being said, I doubt Bitcoin will be outlawed in the West, and generally agree that it will be around for the long term.

The law seems designed to attract media attention for culture-war enthusiasts, rather than have any practical effect.

It feels like this sort of syntax convention could be used to bolt on all sorts of interesting features to the language. Like, at this point, this is basically a sort of compiler hack to add features rather than really do any typechecking.

You have to be a good talker. If you are not you will struggle.

Find a friend who is willing to pay you, or consult for low-pay/free to build up a client portfolio.

Then try to get referrals. The hard part is not understanding new AI stuff, but building the consulting business.

I think it's not the work of updating but trying to remember.

If you are doing consistent frontend you can probably keep it all, but if you only touch it occasionally, it's just too much.

I think a lot of people genuinely believe that AI can/should be regulated.

People at the companies may or may not want to do this for selfish reasons, but regardless, have been basically forced to "ask" for regulation, or face that regulation being applied against them as an existential threat to the business.

Meta seems to have possibly gotten around this through having their model "leaked", but I am not sure what else can be expected at this point. I am hopeful that community efforts will be able to make progress quickly enough so as to make regulatory capture seem more difficult.

Security is often implementation dependent, meaning: you need to understand exactly how your software does what it does before you can understand how to do that securely.

For non-trivial software, this almost always means you will have to build an implementation first, before thinking about how to add security.

Overall, I think security has gotten much better, but for complex software there are always going to be challenges.

This one demands I turn off adblocker, is there a difference in facts?

It's great the team was able to clear him of wrongdoing so quickly!\s

I think that as long as actual human beings continue to contribute content, AI would eventually "refine" that out and incorporate that into newer models. Even if say, 95% of the content in the dataset is AI generated, it will still capture the remainder.

Of course, I don't know if scaling AI with more data will work to improve it or not, and I'm sure that anyone really knows right now.

Congrats to the author! Marginalia is a great service. I hope they find a way to make it viable to keep going, either through donations or some other model.

This is Andy Pavlo, so he probably got sidetracked with https://ottertune.com/

Not sure what op's intention with this was

> Is the only acceptable use of your free time as an adult reading non-fiction about finding the meaning of life?

Of course not! You could also read about productivity. Or maybe a biography of Napoleon? I think those are both fine (got to check with my boss first).

This is not irrational FOMO from investors, it is the reality that to get up and running with a competitively trained model from scratch you need to spend millions. Software startups have traditionally not been very capital intensive, but clearly that doesn't apply to ML startups. Yes, you could have given a smaller seed and asked them to develop a POC, but with their credentials, I don't think their ability to deliver on that is really in doubt. The real question is whether they can create something on the scale of competing with OpenAI. I think that is a pretty good investment thesis, even if it is risky.