The source of about 30% of the hits I took in homework and tests was, to this day, not having memorized the basic values of Sin/Cos(pi, pi/2, pi/4).

Instead of memorizing the values you can learn the shape of the graphs and/or learn the meaning of sin/cos on a right angle triangle.

I think I got ok through a math/comp.sci. degree without memorizing values but I will say the "mechanical" aspects of math including trig functions and things like integrals and derivatives involve drilling those enough to make these things automatic (which is sort of memorizing but not exactly). I was always lazy so I never got really good at that (and my calculus related grades are evidence of that ;) ).

Yeah I pretty much always think through the unit circle to get the value of sin(0) and cos(0) (unless I had been doing a lot of trig lately)

I remember drawing everything out on trig exams though and making myself an on-demand cheatsheet.

Here's an Anki deck for the unit circle: https://ankiweb.net/shared/info/1813562307

Hierarchical model for the win. Consistent with Jeff Hawkins research as well.

I'm not sure why you would think that. There's all kinds of substances that humans take (Alcohol, drugs) that change their mental state. Why is the idea that environmental pollutant could be doing the same, so rejected by people. Is it an ego thing ?

I think people take offence to the suggestion that there might be more autism.

Yes and it tells us zilch about the reality of the situation.

It tells us we do hear a lot more about autism then we used too. It’s a talking point.

True...

Though I think there is more to LSD. It's fair to be skeptical. If the pitch is it's going to fix everything for everyone, or the likes, then it's definitely a lie.

Point being you gotta get your measure to the millions of other clones.

My understanding is they are all still transformers. The tweaks are more about quantization that better to generalize over data more efficiently (so less parameters requires) and improvement of the training data/process itself.

Otherwise I'd like to know specifically whats better/improved between models themselves.

I would join as well. And donate.

This is unacceptable.

Eh? She seems like the first in a very long time to actually be making a legitimate attempt at the job.

Makes her very likable.

If she ends up losing cases to Meta, Microsoft and Amazon, how does that help? She seems to think the chilling effect of the FTC being litigious is victory, but losing every case isn't a great strategy.

FTC lost to Microsoft because the overall case was bewildering. Per Wikipedia, the merger would have made Microsoft the third largest gaming company. Which makes it clear that it faces plenty of competition.

In theory, they could abuse their market position by not selling games for the PlayStation, but you can't sue someone because they might do something. In fact, scrutiny from the FTC and the UK CMA seemed to motivate statements by Microsoft that they would not limit their titles (particularly Call of Duty) to their platforms. These statements were cited by Judge Corley when he denied an injunction against the merger. So even in this case there may have been some benefit.

As for Amazon and Facebook, the cases are ongoing; the former is very recent.

My perspective here is that to do the job was going to be near impossible. No one even knows at this point how to succeed.

And if we need to learn how to succeed, the initial heuristic of applying the law is the only starting point. Failure was the expected initial result.

She can fail for the entirety of her term. We lose nothing from the status quo. But if she can stick to it - then after she's taken enough of the initial beatings, someone else can come and succeed with the learnings she provides.

It's not really "applying the law" if you're repeatedly defeated in court.

This take is so misinformed on how the law works, I am not even sure where to begin. Cases are often decided by judges and how they __interpret__ the law. If you have two generations of the judiciary who have grown up drinking the kool-aid of the magical auto-correcting market, you cannot do shit.

> And yet, this epiphany, that markets are politically structured and don’t have a will of their own, hasn’t made it to one very important place: the judiciary. The same week Sullivan gave his speech, a panel of three D.C. Circuit Court judges struck down a monopolization case against Facebook on the grounds that markets self-correct. "Many innovations may seem anti-competitive at first but turn out to be the opposite,” wrote the panel, “and the market often corrects even those that are anti-competitive." The D.C. Circuit Court panel was bipartisan, and included Republican appointees Karen L. Henderson and Raymond Randolph, as well as Obama appointed judge Robert Wilkins.

> These words undermine Congressional statute, and may devastate the ability to use antitrust law against digital platforms, at least in the D.C. Circuit. The specific procedural question was on the right of state attorneys general to bring an antitrust case over a violation that happened years earlier, as Federal enforcers can. Three judges made a policy decision to disallow that, even as Congress had just passed a law a few months earlier to make it easier for states to participate in antitrust enforcement.

[1] - https://www.thebignewsletter.com/p/all-rise-how-judges-rule-...

The cases are being lost.

It doesn't matter to what degree someone is informed if they're pointing out objective facts.

Indeed, it's actually worse than not doing anything at all, since the record now has precedents that influence future actions.

She's trying to enforce laws that don't exist but that she would prefer did. That's not going to go well, and the real solution is to get new laws passed, not to try to force tortured interpretations of the existing and relatively weak ones...

She also doesn't appear to actually understand the startup ecosystem at all, if she thinks that blindly opposing all M&A is going to help it.

Ah good reminder why I'm paying for the intercept.