it_citizen

« We can’t really make models bigger than 220B parameters »

Can someone explains why?

I was often asked if I wasn't afraid of having kids because of the horrible things awaiting them.

But when was the right time in the history of humanity to not be afraid to have kids if not in the western world today?

Our civilisation might be reaching the end of its course. But even if it ends up tragically, my children still have a better shot at a nicer life than 95% of humans that ever lived.

Try virologist 3 years ago.

The truth is that those 13 years with a stupid high salary will change your relation to money.

You will live in a HCOL area, be around people with much more money, start thinking it is normal to spend 3M on a house, buy stupid things because it doesn't really impact your savings, making so much money will become a source of pride and make you feel good about yourself, and suddenly 2M will seem barely enough to keep up with the social circle you are now part of.

You won't quit and keep chasing the next dollar.

Just personal experience. But I have seen that pattern multiple times.

If money is by far your top priority, sure. But the shortest path to the title might be a miserable one filled with politics, stressful high visibility projects with ton of micro management and extra hours.

In the meantime, I went from junior to principal on a slightly longer path but I side stepped most of the downsides of the shortest path.

I would even argue that because I was less pressured and overworked, I had the energy and excitement necessary to get involved with more initiatives and was perceived as more productive than some of my peers who were powering through massive features!

You can think of it as the 80/20 rule or "Slow and steady win the race."

Aim to run one piece of the company well enough that nobody else need to worry about it. And don't let anyone forget about it.

This will give you more satisfaction and better leverage than a title when it comes to negotiate a raise or request anything else.

I am going to lose some karma over that comment:

As someone who has not watched star wars before being adult but is pretty fond of fantasy and sci-fi, I am wondering if the simplest explanation for the constant disappointments is that the universe and source material are not that great in the first place. Is it possible that in 2023, a good chunk of the love for star war movies is due to fond childhood/90's memories that are impossible to recapture? I am hard pressed to find people that became passionated about star wars by discovering them late in their life.

If I am right, then what to do? People are going to be unhappy if the franchise tries to modernize itself and make drastic changes, but they will also be unhappy if they get more of the same because it won't have the madeleine effect.

They announced a product that is a year away and dedicated almost an hour to it, stretching the entire conference to two hours.

They are definitely not burying it.

Not sure, this is a fiction.

But if I took a guess: more internalization of climate change by the general population, earlier investments into alternative sources of energies, maybe baking in carbon pricing early on before we all became completely dependent on oil.

Whatever the real intentions of Sam Altman are, I think what he is doing might end up being a net positive.

Imagine if fossil fuel companies had helped, early-on, to hammer into the public perception of oil how it might end up being disastrous for humanity instead of burying the studies they made.

I don't think we would have stopped using fossil fuels, but maybe we would have been in a better spot today.

> The fact that scaling continues to work has significant implications for the timelines of AGI development. The scaling hypothesis is the idea that we may have most of the pieces in place needed to build AGI and that most of the remaining work will be taking existing methods and scaling them up.

Is this a pretty consensual viewpoint? Or is it hype from OpenAI?

Comparing the revenues of tech companies to the gdp of a country, even to give a sense of scale is comparing apples to oranges.

Even if a bit unlikely, I would not be completely surprised if the service industry as a whole produced twice the value its produces today thanks to AI in the next 20/30 years. Not to mention the productivity gains in other sectors.

I see a lot of flagged comments of people claiming that "degenerate books" were pushed on people.

Is that really a thing? Is it more than isolated incidents? From my information bubble, those claims seem so foreign to me.

Can someone give me multiple examples with: 1) name of the book 2) problematic example in the book 3) the age of people to which the book was accessible

Maybe we should start with that to explain intersectionality in southern states. :)

If the whole covid vaccine debate taught me one thing is that there are idiots in every jobs, even the most qualified.

Education quality, political polarisation, inequality, poverty index, happiness.

Yet I see carbon capture being a top pick in discussions about effective altruism and by the clean air task force.

I believe their reasoning is that for the amount of money invested, it remains one of the most efficient investment to decarbonize.

I don't know what to do with that information.

I agree. But also I am secretly enjoying that for once, it is Europe getting a product first ;)

Do Canada next!

If they still have enough water that is. Spain might have some thoughts on that.

Good point. I think both?

Nations are not single minded. Political parties, individuals, companies, diplomats, each might have their own particular interest.

But I remember not long ago, "trade and free market will bring democratic institutions" being a pretty popular opinion.

I am glad Germany didn't stop trading with the US after they invaded Irak by producing fake witnesses, or the second time by fabricating proofs, or Afghanistan.

Diplomacy is not black and white.

This is very simplified.

A lot of western countries, Germany first, believed that if we traded and opened up to those countries, they would naturally tend to adopt western values.

It turned out to be incorrect. China being an even better example. But it was not a crazy assumption at the time, and coming after the fact is hindsight 20/20.

Honestly, even knowing how things turned out, I am proud that the west tried.

What does "working as intended" means?

Some sibling comments challenge that it doesn't help reduce the amount of private data stored.

I would argue that GDPR is working as intended on the economic war front. The US weaponized their extraterritorial anti-corruption laws, China is also building similar legal weapons, GDPR is part of the european answer to that: a framework to catch up with other super powers who can pressure foreign companies whenever it pleases them.

After years of development, RSpec is what made me stop testing the code manually. Rspec is one of those things that made me a better developer.

Is that a risk? I don’t know much about warheads but is it realistic that nuclear warheads just detonate along the way? To the point where leaders might worry about it?

Fun fact: it is said that Putin never uses the internet.

I understand the dilemma, but should we forbid ourselves to discuss some topics?

It kind of remember me how the possibility of covid lab leak got discredited early on in the Lancet without any proof and then backfired spectacularly, creating doubts about the scientific community in the population.

I don’t think thought censorship ever helps, we have to push for nuanced views.

Way to twist things around. The article does not suggest anything like that.

Seems to me you are reading about a pretty factual article about the cost of maintenance of nukes but interpreting it as « let’s declare war to Russia, their arsenal probably doesn’t work ».

Because Putin might see it and nuke us to prove us wrong?

I don't see why it would be reckless to consider the possibility of the poor state of the russian nuclear arsenal. The article even points that the tactical or battlefield nuclear weapons are probably just fine.

It doesn't mean that the west is suddenly going to suddenly going to act as if russia didn't have nuclear weapons.