conjectures

Were they autonomous?

Considering it's about not venturing from your hotel room, I think the author does about as well as is possible. But yeah.

What's your point?

So, now you've had a chance to 'get more and more specific' (sic.), this is my understanding of your understanding of an optimal food production policy:

1. Vegan food production wherever possible. 2. Non-vegan food where vegan food is not possible (the Norwegian uplands case).

Is that correct? Or would you like to finesse things any more?

> lots of land.

How much in comparison to farming generally?

You're attempting a bait and switch: 1) Claim veganism is bad because it would reduce output if everyone switched now. 2) Claim veganism is bad because it would reduce output at the margin of an optimal policy.

These two things are quite different scenarios. Scenario 2) is tacitly accepting that most people would be vegan most of the time. But hey, if you want to argue that under both our optimal policies there would be many more plant eaters, fine by me.

I hear what you're saying. The difficulty is that you're another non quant trying to make a general argument and failing.

I'll spell out what a strong argument for the point you're trying to make might look like:

A) An estimate of the total number of calories coming from purely grazed meat on this sort of land.

B) An estimate of the total number of calories of meat consumed.

C) A/B > 3/4.

Do you know what fraction of world calories that type of land is?

Otherwise this objection is a case of 'an edge case exists, and I would like you to imagine that it is the major determining factor.'

Yup. Think I've heard them all over the course of 20 years.

The obvious explanation is that people don't like the implication that their behaviour is unethical and look for clever outs. Rather than simply owning that benefits > costs from their point of view (a la 'I like bacon too much').

Of course questioning motivations is not a point to be made in an argument. However it is an entirely satisfactory Bayesian explanation of why so many people take the other side of the argument with such force.

Also c.f. Venice.

Carrots are vegetables. Vegetables are edible. Is this a picture of something edible? Having a bunch of propositions in a database is not much help if you can't relate them to messy sensory input. It turns out that making sense out of mess was a hard problem. Eventually, logic* will have to be brought in again, but at a higher level, to help out.

* Ofc Hinton's approach is based on logic too, but in the form of probability, calculus etc.

Yup (or at least, some reason could always be found). But they don't shoot you as much. So it's swings and roundabouts really.

> "How do I capture this extremely complicated real world semantics in code?" and languages can't really solve that

This isn't true. It's possible to write a symbolic differentiation system using, primarily, Scala's type system. In an elegant fashion.

Similarly, Julia represents math better than, say python or Java. A vector type should not need to know about inner products of vectors, because the math definition of a vector space does not.

This will have large effects on how easy it is to decompose your code into neat little pieces.

"PhDs can't code" what a load of crap.

I don't think this is likely to happen. There is almost surely a class of interesting questions it takes AGI to answer but a math PhD to pose.

In what sense?

FYI people design trials with breakpoints, bandits etc. The issue isn't early stopping, the issue is inappropriate analysis of early stopping.

The difference is that your average ML marketer can't plausibly say their stuff is just too cool to even tell you what it does.

C.f. the GPT2 announcement which got some pushback, and that was with plenty of concrete examples.

Everyone involved in this article has an incentive to overstate its abilities. The creators will say it is better than it is to get more work. Defense to create a deterrent. Journalists to get clicks.

Given the baseline for government understanding of AI is poor, my prior on how impressive this thing is in reality (as some sort of AGI pathway breakthrough) is pretty low.

I would bet they have some large database and some good, but mostly conventional IT around it.

TL&DR: a few vegans became meat hipsters. The author thinks being a meat hipster is a very good thing.

This is now a little out of date given 1.0+ has been released. Still a decent article though.

Rather than the story of financial degregulation?

Sceptical.

This is a question I ask.

The UK policy was, effectively, 'paaarty!'

Yeah this article isn't that clear so don't worry about it.

For example, to suppose that particle filtering is an extension of Bayesian probability is a category error. If this just stemmed from communication difficulties, it still matters from a reader perspective.

What does it mean to be 'poorer in a monetary sense' if it leads to comparisons where person A has better healthcare, transportation, information access etc than person B but B > A?

Your comment focuses on standard of living issues. But your earlier claim was that these are separate from 'monetary wealth'. I think you would need to specify what 'monetary wealth' is separate from 'standard of living' in order to make a convincing argument.

You'll find this hard to do, because any wealth measurement is either a thing, or an instrument that will get you a fraction of the things other people have. Financial wealth is cashed out in stuff. So cross temporal comparisons are going to involve comparing the stuff two people can command. And the stuff they can command is embedded in the technology and society of their respective times.

If you think this is a conflation, perhaps you could explain how and why they shouldn't be?

Inflation figures over 200 years are pretty dodgy. How much was a dose of penicillin, an iphone or a flight to New York in 1819? There's a very real sense in which very average Westerners now are far wealthier than Darcy.

Conversely Darcy would have had far more power than the someone with net worth £16.8-28m would today. He'd probably have had a large say in who the local MP was, the local priest. Probably could turf a large chunk of the local population out of jobs and homes. Likely had a small army of servants.

Great to see this finally land - thanks for all the team's work. Looking forward to giving it a whirl.

Threading is something of a prerequisite for acceptance as a serious language among many folks. So great to not just check that box, but to stick the pen right through it.

The devil is always in the details, but from the doc the interface looks pretty nice (given threaded code generally looks like it was written by Cthulhu).

Smart insulin pump sounds awesome. Bring them on. If it really is a smart insulin pump. Trouble was the article author was talking about 'reading his emotions' etc. There's a red flag for bullshit. AFAIK no publicly disclosed technology can, in general, read emotions. Also, why whatever signal they are using as 'emotions' is a perfect predictor of insulin levels, I don't know.

It's like every other get rich quick scheme: as soon as you've heard of it chances are it's too late.

> When he was diagnosed with type 1 diabetes, I hacked his insulin pump and built an AI that learned to match his insulin to his emotions and activities.

Wow, hope there're no bugs in that one. I'd hate for the kid to be weeping over Titanic and get od'd with insulin.