sokoloff

You already have the choice to cancel your subscriptions and run local LLMs that keep your data local and private to you.

Other people also have that choice and many choose differently from you, but no one’s taken your choice away.

Then parents who do morning routine or drop off have to come in an hour later, and their partner also wants to wake an hour later so they can stay up with the family an hour later, which means they don’t want to open the coffee shop until an hour later, so now the random commuter wants to start later so their coffee shop is open, and pretty soon it’s just as easy to say “ah screw it, we’re just going to change the clocks twice a year…”

I suspect the one who fires 80, does twice as much, and (maybe optionally) pays the 20 125% (as if they were 25) probably handily beats them both.

I see tons of straight-up fully AI videos being offered, whether it’s NotebookLM discussions of some topic, some nonsense reading of “accepted answers” to a tech query, or the maddening “here’s three ways old money acts”, I assume based on being interested in tech topics and retirement saving topics.

Taking the low end of “more than $100K” as being exactly $100K, 3.2M people earning $100K/yr each is $320B/yr and many earn way more than $100K, obviously.

I think either your $50B/yr in the US figure is low (would mean less than 1 in 6 dollars is earned in the US) or, I think more likely, your 3.2M figure is too high.

Where did you source the 3.2M figure?

It’s confirmation bias; you notice and remember the misses while mostly overlooking the hits.

I think it will create a lot of jobs (overall) that pay twice the median wage.

The fact that some of those jobs would have paid 3-5x the median wage in 2021 isn’t a scathing indictment of AI, even if it’s inconvenient for the slice of the world who could have made 4x the median wage and now only makes double it.

Then you must think it exists outside the human brain, right?

> it is helping people realize their ideas are stupid or their busywork doesn’t matter way faster.

While this is a possible explanation, it seems far more likely that people are finding more private/solo ways to pursue whatever their equivalent of “make the next Facebook” or “predict the stock market”.

I would be very surprised if the was a reduction in total duration of pursuit of these type of ideas. I think there’s just a reduction in them asking you as a dev to help.

I agree that it’s a tool to explore risk. I just think it’s more useful if you pull the likely universe of outcomes from the sequences of the past (sliding your retirement across multiple starting dates) more than worry about what happens if we get multiple -43% years in a row. There’s no realistic portfolio that will survive that, but also no realistic underlying economic model where that happens (or if it does, where what you have saved really matters).

“I just need to find another 12K of conventional memory to launch this game…”

I don’t think a pure Monte Carlo is as helpful as “replaying” past actual scenarios (or a hybrid of the two).

I don’t have reason to believe that 40 years in a row would all have randomly and independently chosen results from a distribution (pure Monte Carlo) as a better model than past actual results.

I think most everyone but you interprets “GPS radio” in GP’s post to mean a “radio capable of reviving the GPS signal” same as I would call the thing in the middle of car’s dash an AM/FM radio.

Who the hell designs a scale that can’t work in a typical bathroom?

In what room did they possibly imagine their device would be used?

My point is that LLMs will soon be seen as “things you just have” not as "solely for the purpose of the interview".

I think that's probably already the case for the median software engineer and maybe even for the median software engineering candidate.

The relevant question is “how incidental/baseline to life is this expense?”

We require clothes for on-site interviews. We require phone service for phone screens and electricity and internet service for video interviews. We don’t pay gas/mileage for on-site interviews for local candidates. People think of those as incidental/de minimus expenses and we ignore them.

I think LLM subs are going to be just as baseline as mobile phone service very soon (far, far faster than mobile phone adoption was in the US or Europe).

It’s not all candidates, only candidates who progress to a certain gate in the process.

I think “we offer a $50 (or $20 or whatever) Visa/Amex gift card as part of this interview step” is easier to admin, might skate by finance more easily, and avoids all the tech support angles of “here’s how to use our specific tooling” and concerns of them misusing your key for something nefarious.

The highest production year was 1943, with 65 completions (including those built for Britain) and of those 65, only 6 were Essex (fleet) and 9 Independence (light) carriers with the other 50 being small escort carriers.

> Nobody would chose to sacrifice themselves and their descendants (because that's what such an expedition is about), for mere curiosity.

I bet if you had a one-way mission to Mars leaving in 12 months, you’d have a flood of applicants. Many hundreds minimum and probably several thousand.

Out of 8 billion people, assuming you had selection criteria that were even as strict as 1 in 1000 qualify, out of that 8 million, I think you’d easily be able to raise a crew.

I would expect indoctrination something like what keeps alignment among devout religious practitioners would be forced to emerge. The alternative is extinction on a spaceship hurtling into the vastness of space.

I suspect education on that ship would look a lot different than it does on Earth.

> the Apollo program supported 5-digits worth of very good jobs nationwide.

It was 5-figures of direct NASA employees and over 350K workers in contractors.

They have dozens of offices in the UK and Germany, 13 in Netherlands, …

It hardly seems surprising (nor damning) to have one in Cyprus.

Isn’t basically every typical residential lease advertised exactly that way?

Neither QQQ nor SPY is a leveraged fund.

But you can fund research in other labs/companies and then bring in-house the research you like to commercialize and leave research results you don’t like with companies like “Exxoff Research 37” if having negative research results becomes a liability.

You can then even truthfully claim to be funding basic research that doesn’t directly benefit your commercial interests.

Is what a scientist at Exxon wrote in 1988 more relevant than what hundreds of fossil fuel extraction, refining, transportation, and retailing companies are doing today and tomorrow?

An existential threat to human survival demands action to stop all of them not selectively or preferentially only the ones who knew a little more than others 38 years ago.

Doing the latter is not only nearly totally ineffective (as they’d be instantly replaced by others taking market share) but also ensures that companies take great pains to never do in-house research on things that might be harmful.

I suspect that 38 years from now, future versions of us will look negatively upon Greenpeace, Friends of the Earth, and Sierra Club's opposition to fission reactors for electrical power generation, whether as a transitional approach to reducing fossil fuel combustion or longer-term base load generation. (Some current versions of us are looking negatively at that, but not all.)

I don’t think selling street crack is a good thing and oppose everyone who is selling it (and to your last point in your upthread post, I wouldn’t worry about the supply chain players, because “who works to supply crack and doesn’t know that crack is probably not great for you?”)

I don’t pick and choose to punish one of the dealers more harshly than the others because that one dealer had some internal research that cast doubt as to the likelihood that he would, at some time in the future, achieve a workable solution to a healthier/safer crack substitute that he currently publicly claimed to be working on.

People seem to be conflating “this company didn’t publish everything it thought about everything” (a standard we don’t hold companies to in general) with “…and the topic was climate change, so off with the company…”

Would you also like to know if there was an EV company or environmental group financed or sponsored astroturf campaign against ICE vehicles?

Frankly, I hope that both of those things happened, so that people could make their own minds up after an adversarial process. (We have one pure-EV and one ICE car (non-plug-in hybrid, but that's just ICE with extra equipment for marginal efficiency gains), because the cars have different use cases that they're good at.)

It’s not like you can squash only PR campaigns that you happen to be misaligned with, so If you want your own biases to be able to communicated, you have to let others do the same.

That’s seems ass-backwards though. If you want to punish Exxon because they contributed to global warming by satisfying the demand of billions of people for the energy embodied in fossil fuels, it seems like you should directly attack the thing you don’t want: extraction and combustion of fossil fuels.

Don’t kill Exxon and let Shell operate.