While one can reasonably argue if depreciation should be 3 or 6 years, the article is utterly dumb, quote:
> "However, some types of assets fall in value more sharply early on, then stabilize and decline more gradually on a predictable curve. A case in point: According to Silicon Data, which tracks pricing for Nvidia chips, the average resale value for an H100 system in its third year of use recently was around 45% of the price for a new H100."

1) Depreciation is unrelated to market value. Real estate is depreciated over 30 years, whether or not the value is increased or decreased.
Useful life is however long the item is actually used and contributes to company P&L -- unless otherwise declared so by accounting guidelines. There are no binding guidelines here.

2) Annual depreciation _could_ be taken (but usually isn't) as "book value - salvage value" / useful life. Using the numbers in article, if 2.5 years in, value of asset is 45% of purchase price, annual depreciation is ~55%/2.5 = 22% -- resulting in exactly the same depreciation as a 5 year accounting schedule.

Who would have thought...

This has been going on since crapto became big enough and its likely a main reason crapto is still around? Crime-support in the from of tax evasion, crime financing and money-laundering was always a major application scenario for crapto. Obviously, it also serves as a scam vessel by "value" manipulation (see Musk and Trump, for examples doing that).

This is called "being careful". They could just have done what Boeing does and risked a few 100 dead but avoided that costly "recall". Instead they determined the possible causes and eliminated the most likely ones, and those include an unknown software fault. They currently are not finding that fault and hence they think it may have been a rare but possible event like a bit flip.

Time for the US to nationalise all things vehicle. Registration and taxes. Emissions and smog checks. Safety inspections. Dealership laws and regulations. Driver licensing (including for trucks, busses etc). Road rules. The lot.

Fuck that.

I want the govt more OUT of my life, I dont want to give them more pathways into my life....

Nice! Also pretty accurate. Well, build houses of cards on sand and this happens. What I do not get is that these people do not see it. It is neither difficult to understand not is it without precedent. In fact, there is a very large body of examples from when other engineering disciplines struggled to get to maturity. And it is even in the mainstream media (for example the Titanic, Tchernobyl and Fukushima).

These people must be determinedly dumb, uneducated and incompetent.

Indeed. The problem of IT security has always been asymmetric, but AI makes this massively worse. The small advantages AI offers on the defenders gets annihilated and then steamrollered over by the advantages to the attackers.

Without the influx of tons of Windows people (that do not get it) into the Linux space, systemd would never have been a thing. That same problem could or could not happen with the xBSDs, but it should at the very least be far away. Meanwhile, all my Devuan installations and the few remaining non-systemd Debian installations continue to run perfectly fine and with no gross security problems.

I am currently beginning to think that Rust may actually improve software security. Not because of its features, but because of that steep learning curve. We have too many incompetents writing software. Rust may be too hard for them. If so, good.

But please, get the Rust spec done. A "secure" programming language without a spec is simply embarrassing.

Yep. The nuclear industry is famous for its bombastic empty promises. And the morons believing them.

And then run it without incident and good uptimes for at the very least 5 years. At that time I will start listening.

Before? The whole thing is just fantasies and hopes and dreams, no substance.

My prediction is that all this "AI Agent" stuff will be dead within at most 2 years due to spectacular failures. They already have created really easy to use attack paths on people's systems and data several times now and they cannot really secure them.

I find it totally fascinating how determinedly these "decision makers" try to ignore that LLMs cannot deliver anything but a tiny fraction of the claims made about them.
Groupthink, FOMO, plain old stupid, all intensely at work.

In other news, LLMs and generative AI lose potential application scenarios that would make operating them worthwhile left and right. Apparently different art communities see it as fraud if you use generative AI and not declare that. And if you declare, nobody wants your stuff anymore. Nice!

The co-generation idea has been around for decades. Although there are some successful implementations, there are many problems to surmount to make it practical and profitable.

One of them is energy output efficiency for the lifecycle of the heat source, customers to buy the energy output, competition from other energy sources (like artificially low petro-energy costs), and sufficiently cheap capital to make the return on investment work over the projected life.

Didn't happen before, and it's unlikely to happen now, despite the cool-factor of the idea.

... you have to drink the KoolAid.

Will his co-CEO AI Teammate take his job?

Will Amazon's Teammates effectively infect the other AI Teammates to recommend Amazon's products, delivered by robots, for gear made by robots, to robot purchasers?

Will all the AI and robotry be able to form a new society without those pesky humans-- now unemployed? We wait with baited breath for our new AI overlords.