But how much effort would that require? ... is what I was going to say but in this age of "claude reverse engineer this doohickey" maybe the better question is how much compute and how many tokens it would require.
In addition to that birds don't tend to emit much RF. And spectroscopy of the visible light from the materials comprising the body should be quite different unless the drones are covered in harvested genuine feathers.
Probably, but only due to sheer laziness. It should be easy to distinguish electronics from birds at sufficiently close range. Plus I'm certain the physical behavior doesn't hold up to close scrutiny.
You don't need clean data to train, just "reasonable" data. General bulk data from the wider world isn't the difficult part. AFAIU task specific RL environments, custom annotated datasets, and access to the raw compute necessary to actually perform the training are the primary limiting factors. The latter in particular is no joke. Most companies won't be able to afford the pretraining for a frontier sized model. Even "small" models can run into the millions USD IIUC.
> The GPUs are probably 80% of the total cost or more. And they probably only have a 3-5 year primary lifecycle.
There's plenty of data indicating that inference at typical API prices is profitable after accounting for the cost of GPUs. However that doesn't account for the arms race of training new models while performing R&D.
It isn't whataboutism if it's directly relevant to the point at hand (as opposed to attempting to distract from it). The question posed effectively amounted to "why is this particularly noteworthy compared to everything else of a vaguely similar nature that's been going on".
Meanwhile intel (as but one prominent example) was also researching and manufacturing GPUs for most (all?) of that timeline yet somehow managed to fumble it. Quite often IMO simply not screwing up while already holding all of the necessary cards is a lot harder than it looks.
I certainly don't condone it but nonetheless the underlying motivations as well as the potential consequences of those behaviors seem almost entirely unrelated to me.
That's overly general as by that logic nearly every action anyone takes is equivalent. Reliably show up for work? Self serving pursuit of a paycheck. Grocery shopping? Feeding oneself is also self serving.
Someone selling secrets (for example) I would generally assume to be motivated by pure greed coupled with a fairly blatant disregard for both criminal and financial risk.
In contrast, I understand affairs to commonly be motivated by any number of things ranging from lust, various relationship problems, the thrill of something taboo, etc. The risk is also typically more social than financial and certainly not criminal in nature (at least in the west).
> That's overly general as by that logic nearly every action anyone takes is equivalent. Reliably show up for work? Self serving pursuit of a paycheck. Grocery shopping? Feeding oneself is also self serving.
None of those examples involve breaching a trust or lying in order to pursue that self-serving goal. I didn’t say that having a self-serving motivation was a problem by itself. The problem, with both cheating on your spouse and cheating your company, someone is putting their self-serving interest ahead of their commitments and duties to other people and lying in order to avoid consequences.
> Its an attempt to retcon history to vilify the US.
An attempt to vilify ... the side that eagerly deployed nuclear weapons on civilian targets? I'm from the US and I'm not someone who is blind to pragmatism but your argument here just seems completely absurd to me. We clearly committed an atrocity with somewhere between minimal to no legitimate justification.
Then again, after seeing the adjacent comment about avoiding soviet involvement it occurs to me that perhaps we inadvertently did them a favor in a roundabout way. Japan probably would have lost a lot more than the kuril islands otherwise.
Where did you get the idea that the USA was “eager” to deploy the atomic bomb?
>Stimson wrote in his diary afterwards that after his briefing, Truman felt “the terrible responsibility that such destruction placed upon us here and himself.” [1]
[1] “Memorandum of Conference with the President, August 8, 1945 at 10:45am,” in Stimson diary (8 August 1945), Henry Lewis Stimson Papers (MS 465), Manuscripts and Archives, Yale University Library.
For starters, how many days did it take before we dropped the second one again? I don't think it matters what any of the participants wrote or said given the events that we know to have transpired. There were clearly many other paths available to take.
Easy to say 80 years later, not having any of the personal experience, and handwaving away first person accounts.
You can certainly choose to believe that the decision makers were eager to use nukes and didn’t consider other paths, but you have to acknowledge that opinion is coming entirely from your own intuition and intentionally excludes all contrary evidence.
It's literally in Truman's diary. [1] I mean I suppose he doesn't "literally" say 'I'm anxious to nuke the Japs', but he comes pretty close:
"Discussed Manhattan (it is a success). Decided to tell Stalin about it. Stalin had told P.M. of telegram from Jap Emperor [actually from the Foreign Secretary] asking for peace. Stalin also read his answer to me. It was satisfactory. Believe Japs will fold up before Russia comes in. I am sure they will when Manhattan [A-bomb] appears over their homeland. I shall inform Stalin about it at an opportune time."
And on top of this, the Japanese had been negotiating their surrender with the ostensibly neutral USSR hoping to get more favorable conditions with their assistance. And the US had been secretly intercepting these cables and knew all about it. [2] And on top of this the US had also already created another secret agreement with the USSR such that they would declare war on Japan on August 8th, which would end any hope at all for Japan. We nuked them on August 6th, and then again on the 9th.
Not quite. The main negotiator they were using in the Soviet Union was fully aware of the reality of what was happening, long before the nukes and Soviet declaration of war. He was suggesting unconditional surrender except with the promise of the continuity of the Emperor, which is exactly what they ended up going for.
The problem is that he did not, independently, have the authority to make such an offer. And in the mainland things were also confused. The Emperor was treated as a borderline divine figure, but that also meant he was mostly hands off except as a lost resort. He was pushing for surrender behind the scenes, but the war council was the one making the decisions of what would be offered, and they were completely divided. Half wanted to fight, and half wanted peace.
This division continued even after the nukes and Soviet invasion. What changed is they chose to allow the Emperor to make the final call, and he chose to surrender as he had been angling for for quite some time already. The nukes played little role in this decision. The pro-war Japanese correctly believed that we couldn't have many such weapons and were, in any case, prepared to fight to the last Japanese because they thought it'd be glorious. The thing that did change their mind was the Soviets entering the war meaning they not only no longer had any intermediary with with which to hope for more favorable terms, but also were now facing a massive invasion from the north as well.
The diary is full of revealing quotes like "Discussed Manhattan (it is a success). Decided to tell Stalin about it. Stalin had told P.M. of telegram from Jap Emperor [actually from the Foreign Secretary] asking for peace. Stalin also read his answer to me. It was satisfactory. Believe Japs will fold up before Russia comes in. I am sure they will when Manhattan [A-bomb] appears over their homeland. I shall inform Stalin about it at an opportune time."
For some timeline the US had secretly arranged for Russia to declare war on Japan on August 8th. We nuked them first on August 6th, and then again on the 9th. The goal was to get them to surrender to the US before the Soviets made significant inroads in Japan.
Do you think what they reveal is the truth? Was the diary meant to be published and become "a source"? Was Truman informed about everything important going on in the US? About the reality (the truth) of the Manhattan Project? Did he even need to know? Wasn't it preferable for him to simply believe in order to play his role?
Just some questions to ponder as to the quality of this source and what can be deduced from it.
> the side that eagerly deployed nuclear weapons on civilian targets
No one was eager. This sort of lie destroys discussion. This site is only about discussion, so please don't keep trying to destroy the only thing the site is about.
This is so profoundly and comprehensively ignorant about pretty much everything about the pacific theater of WWII that it's hard to know where to even start. I would suggest you start reading about the conduct of the Imperial Japanese Army in China, Korea, Philippines, and pretty much everywhere else they invaded and occupied, much of it before the US even started to consider entering the war.
reply