If I don't trust my doctor, I can ignore them and find a new doctor.

If I don't trust my therapist, I can ignore them and find a new therapist.

If I don't trust my housekeeper, I can fire them and hire a new housekeeper.

If I don't trust a police officer, it doesn't matter. If they detain me and order me to step out of my vehicle, I have to comply under threat of the law and violence. I don't get to only listen to police officers whom I trust.

That is why they must be held to a higher standard, because they wield elevated authority not granted to ordinary citizens.

A police officer who has demonstrated such a reckless disregard for the law and safety can not be trusted as a police officer to uphold the law.

Their point was pretty clear: you can’t opt out of a cop you don’t trust. You can opt out of seeing the unhealthy doctor.

It was, and I was pretty clear that I understood that. It dodges my point, though, and I've asked that they acknowledge it: should you distrust unhealthy doctors, not can you? Is unhealthiness in one's personal life disqualifying for the ability to provide stellar health advice in one's professional life? Should cops be held to the standard of being exemplary citizens who don't even speed? Have you ever sped? Do you know anyone who has never sped?

> should you distrust unhealthy doctors?

Maybe! Is their health status directly related to their specialty? Is it a readily curable condition? Is their advice reasonable?

Or are they a lung cancer specialist chain smoking cigarettes at the appointment?

> Should cops be held to the standard of being exemplary citizens who don't even speed?

Yes, cops should be held to a higher standard than the general public. Being a cop while committing a crime should be an aggravating circumstance in the justice system, not a get-out-of-jail card.

Should we expect perfection? No. But 547 speeding tickets is unacceptable.

Yes, the doctor's advice is reasonable and no the doctor is not smoking at a lung cancer appointment. The premise is that they are messing up off the job. If you think an oncologist shouldn't be able to get as good of a job because they smoke cigarettes or eat burgers, that is where we disagree. Apart from calling that illiberal or saying it has negative utility in its consequences, I don't know how to argue that; it's a values difference. I appreciate you actually taking the position, though.

Same for your cop positions. You say they should, I say they shouldn't. If it's clarifying, I can add that I agree that cops should be held to a higher standard while being cops, ie. that things like qualified immunity are working in the wrong direction, and that they shouldn't be held to a lower standard, on- or off-duty.

As far as I'm concerned, speeding tickets in the course of your private life are between you and the ticketing authority. If he's not paying his fines, if he's violating the social contract, sure, escalate. If we want to punish speeders with more than fines because of endangerment, like the article strongly suggests, sure, change the law. But as long as he's compliant with his fines and we're only giving him fines, it's not just to continue to pile on consequences.

Each of those examples varies widely, and I don't think most people would treat each of them the same way.

In general when the stakes are higher and the ambiguity of outcome is less clear, secondary signals become more important.

Concretely: I don't give a shit if my housecleaner doesn't make their own bed as long as they make mine; the outcome I need is easy to verify and the stakes are fairly low so the secondary signal doesn't matter very much. Conversely, I care a lot if the therapist I'm relying on to help me manage my depression is visibly unable to manage their own; the outcome I need has a slow feedback loop and the stakes are high so I'm much more likely to rely on secondary signals like "is this person able to manage their own mood successfully?"

In your version of the therapist example, you don't trust them to do their job because they are failing to do their job. This is fine by me. My issue is with punishing people at their job over actions taken outside of their job, as in the example under discussion.

> I don't understand the basis for applying this standard to cops.

Because those other examples don't involve breaking the law.

At the risk of pointing out the obvious: society holds breaking the law to be more serious than being fat or not making your bed.

Now, speeding is very much a lesser form of breaking the law...but then again, very few people have literally hundreds of speeding tickets.

Huh? Your examples are all people failing at something that merely resembles their job. For the analogy to work, Giovansanti would need to be failing at something in his personal life that resembles traffic enforcement. Instead he's doing the exact thing his job exists to prevent.

And not also the people who think that acquisitions are being made primarily because colluding CEOs have some interest in creating the illusion of a healthy economy?

Taking my paycheck and multiplying would be an excellent measure of my yearly salary. I don't understand how that analogy is meant to imply that the approach is nonsensical.

because your paycheck comes only one day per month, so that estimate would bee too high by a factor of 30

But then the analogy doesn't work at all? ARR isn't calculated by multiplying revenue to 30 years and then saying that that's 1 year's revenue; what criticism is being made here?

they pick some month with the higest revenue. Unlike your income, a business makes different amounts each month based on some trends and many other factors, and they can varry wildly.

You'd want to compare against the fraction of training attributable to the image

This is a plainly dishonest comparison. A single H200 does not need to run continuously for you to generate a dozen pictures. And then you immediately pivot to comparing the paint usage against "the grand scheme of things"- 700W is nothing in the grand scheme of things.

In fact it's pretty fair.

Many people think that when a piece of hardware is idle, its power consumption becomes irrelevant, and that's true for home appliances and personal computers.

However, the picture is pretty different for datacenter hardware.

Looking now, an idle V100 (I don't have an idle H200 at hand) uses 40 watts, at minimum. That's more than TDP of many, modern consumer laptops and systems. A MacBook Air uses 35W power supply to charge itself, and it charges pretty quickly even if it's under relatively high stress.

I want to clarify some more things. A modern GPU server houses 4-8 high end GPUs. This means 3KW to 5KW of maximum energy consumption per server. A single rack goes well around 75KW-100KW, and you house hundreds of these racks. So, we're talking about megawatts of energy consumption. CERN's main power line on the Swiss side had a capacity around 10MW, to put things in perspective.

Let's assume an H200 uses 60W energy when it's idle. This means ~500W of wasted energy per server for sitting around. If a complete rack is idle, it's 10KW. So you're wasting energy consumption of 3-5 houses just by sitting and doing nothing.

This computation only thinks about the GPU. Server hardware also adds around 40% to these numbers. Go figure. This is wasting a lot for cat pictures.

And, these "small" numbers add up to a lot.

Definitely worth considering in a world in which there are any H200s idling in data centers.

Now that's one fine No True Scotsman.

    A: GPUs use a lot of power!
    B: Not all of them are running 100% continuously, eh?,
    A: They waste too much power when they're idle, too!
    C: None of the H200s are sitting idle, you knob!

I mean, they are either wasting energy sitting idle or doing barely useful work. I don't know what to say anymore.

We'll cook ourselves, anyway. Why bother? Enjoy the sauna. ¯\_(ツ)_/¯

B is supposed to be me? I said the H200 doesn't need to be running continuously to generate a dozen images. If a million people generate a dozen images, it no longer makes sense to compare to the costs of a single artist for 6 hours. I really don't understand why this is hard and that makes this feel very uncharitable.

I'm not saying that this isn't "true idling", I'm saying that idling H200s simply don't exist, i.e., I disagree with B. Do you, A, even disagree?

> they are either wasting energy sitting idle or doing barely useful work

Now here's a true (inverse) scotsman, or more accurately, a moved goalpost: Work on things you don't deem valuable is basically the same thing as idling?

> We'll cook ourselves, anyway. Why bother? Enjoy the sauna. ¯\_(ツ)_/¯

I'm very concerned about that too, but I don't think we'll avoid the sauna with fatalism or logically unsound appeals to morality about resource consumption.

Only either indirectly or extremely partially. Everybody was driving by the 40s in the US but the obesity epidemic didn't start until the late 70s. Really, obesity is an example of exactly what the GP was talking about- our evolved relationship with food is not inherently good, and it's better for us to change our behaviors than to abandon our advancements and return to the food-scarce world we're adapted to.

>our evolved relationship with food is not inherently good, and it's better for us to change our behaviors than to abandon our advancements and return to the food-scarce world we're adapted to.

So are you arguing we should change our relationship with human intelligence? What does that even mean?

My intent was to argue that the obesity epidemic is not a supporting analogy for their point. I didn't mean to imply that the problems with it analogize back to AI.

I don't think you can use this datapoint for this purpose. Cops are employing the paranoid strategies already, so there's no way to discern between 'these strategies are needless' and 'these strategies are effective'.

You could probably do a comparison between jurisdictions where police homicides are common and jurisdictions where they aren't common. Assuming that there are cultural factors anyway.

Like sure, areas with higher rates of criminal violence will probably have more police homicides, but it's likely enough that you can pair things up based on rates of criminal violence.

Yeah, call me when Yann incorporates the four humors and the elemental force of fire, from which we draw life. Metal lacks the nature for this purpose.

Yeah, but surely this goes in the other direction rather than answering the question; average > median