I suspect the 2007 in the title refers to the fact that bills were passed to ban this stuff in 2007, which is when the PRISM program (also illegal domestic surveillance) got started.
(The title makes it sound like warrantless surveillance lasted from 2001-2007, but I think it means the article only covers that date range.)
This one is unusual in that the government started bailing out the AI companies last year. Usually, it waits until the bubble pops, and then starts the bail outs.
That's standard operating procedure for Trump though.
He did the same thing in 2016-19 with the zero interest rate policy + tax cuts even though the economy was strong. Any macroeconomics book (or NPR station during those years) will tell you that doing that creates short-term economic growth, but sets the next administration up for [hyper-]inflation.
Of course, that happened, and those same books go on to say "and, usually, because inflation takes a bit to kick in, the next president will be blamed. This is why we have an independent Fed".
So, this time around, he's trying to pull the same crap by dismantling the Fed, and, until then, lean hard into deficit spending to keep unemployment low. Last year, money went to data centers, and domestic paramilitary actions and prison build-outs. This year, we have those things and a new pointless forever war.
However, it's not working the same way as it did last time. He's done so much other collateral damage that we're in a "boomcession" where the economic indicators become untethered from reality. So, they show growth, but people's quality of life, spending power, job security, and so on all decrease.
For example, a piece of the GDP is "how much does your bank screw you per year on your checking account?". This is treated like discretionary spending, and it's gone up from a few hundred a year to over $2000 in 2025. That increase counts as economic growth, instead of institutionalized theft.
Medical spending increases drove all the US's GDP growth last quarter. The quarter before that, it was spending on AI datacenters that's backed by junk loans and federal dollars.
Anyway, I don't have an answer for your question better than "bubble", but the current economic cycle is not what you described. It is a "boomcession". As far as I can tell, it's a new class of economic disaster, at least in the US.
There's no legal basis for blocking defense contractors from using them. Trump's claiming he can do so, but the law doesn't back him up. He'll lose in any fair court, or any corrupt court that values billionaire interests over virtue signaling to the orange one (like the Supreme Court).
Also, they got a huge PR win, and jumped to #1 on the Apple App Store. Consumer market share is going to decide which of the AI companies is the market leader, not fickle government contracts.
If you look at what generates cash, it's corp to corp. That's across most industries. While there are markets that are consumer mostly, LLMs have immense and enormous business facing revenue potential. The consumer market is a gnat in comparison.
There are always Executive Orders that can enforce that. It is not like in the movies where they will sort stuff out in 2 weeks in a single trial. It is going to take years, and we'll see if Anthropic survives that.
They basically are cancelling the contract, but there are some nuances on Anthropic's side. The contract probably has stipulations that prevent them from doing it overnight, so it might be illegal (but ethical) for them to just turn off the API keys.
Also, doing that might have bad second order effects with bad ethical implications.
For example, when Musk decided to pull the plug on a bunch of starlink terminals, he (intentionally and knowingly) blocked a US-funded attack that would have sunk a big chunk of the Russian navy, which certainly prolonged the Ukraine war. That was clearly an act of treason (illegal).
Anyway, just turning off Claude could kill a bunch of civilians in the region or something. It depends on how deeply it's integrated into military logistics at this point.
Anyway, your point certainly holds for OpenAI:
They walked into a "use ChatGPT for war crimes, and illegal domestic surveillance / 'law enforcement'" deal with open eyes, and pretty obviously lied about it while the deal was being signed. I don't see any ethical nuance that would even partially excuse their actions.
Though, I guess IBM did get away with lots of stuff that... Actually, did any supply companies in the WWII German war machine actually get in trouble for war crimes, or did they just go after officers and the people actually working in the camps?
The company selling punchcards that were used for logistics was apparently fine. What about the people making the gas canisters, or supplying plumbing fixtures? The plumbers? Where's the line?
Wondering, since this is increasingly becoming a current events question instead of an academic concern.
There were the so-called Subsequent Nuremberg Trials (12 of them). Among them were the trials of IG Farben (gas chamber supplies, Zyklon B) and Krupp (armament of the German military forces in preparation of an aggressive war)
I'm under no illusion that all the perpetrators of war crimes were held accountable but it's not a bad model.
Yes, Musk is guilty of treason for exactly that reason. He directly sabotaged a major US military operation in Ukraine.
However, the military is bound by US and international law. It's clear they're not going to obey either of those with respect to this contract.
On top of that, Anthropic has correctly pointed out that the use cases Trump was pushing for are well beyond the current capabilities of any of Anthropic models. Misusing their stuff in the way Trump has been (in violation of the contract) is a war crime, because it has already made major mistakes, targeted civilians, etc.
Grok is chosen because Musk spent $250+ million to elect Trump and is expected to underwrite the 2026 elections. Also, a lot of Trumps and their friends are invested in SpaceX. So they give them money too, but use OpenAI or Claude. I have a feeling that the military likes Claude more
They "chose Grok" for political optics, but they don't seriously intend to use it because it's actually just benchmaxxed garbage - hence why they worked with OpenAI.
CLI is probably more reliable. Also, the ergonomics for the person setting up the machine for the AI are better. They can check to see if the command is working without screwing with curl. It's also possible a human might want to use the software / service they're paying for.
Manufacturing tolerance is the term for "how close are they all to being the same shape?" Good tolerances are usually a prerequisite to good build quality, but not always.
For instance, cast iron pans can have poor tolerances (be off by fractions of inches), but, as long as they're not warped, and the metallurgy is solid, they could last centuries, and people would say they have good build quality.
On the other hand, a stainless steel pan that's volumetrically-perfect, but has faulty internal welds on the laminated bottom could fall apart after a few uses due to heat strain snapping the welds. That'd be terrible build quality.
What have they produced recently? I found a few lists online and looked at Wikipedia, and their big hits are all > 10-15 years old (or sequels/re-releases). Many of those are decade-old acquisitions of franchises that were ancient at acquisition.
Distribution and marketing? I wouldn't even know how to buy one of their games. I have a Linux gaming PC, a Mac, a switch, an iPhone, iPad, Apple TV and a XBone. We spend a few hundred dollars a year on video games, but I haven't seen anything suggesting any MS studio products work on any of our hardware, or are available on any distribution channels that reach any of our devices. Maybe they're on iPad, iPhone or Android? I haven't checked because we don't use those for gaming.
Windows is at ~ 95% of steam market share, so I guess that's one bright point for MS studios. However, many game developers release on Steam and console, so it doesn't imply that 95% of those other studios' customers could run a MS studio game.
Customer retention? The last time I plugged the XBone in, I spent 45 minutes screwing with bugs in the account password dialog, finally logged in, and then walked away. I unplugged it for good after the updates completed. In contrast, I spent less time than that installing Linux + Steam on my most recent PC. I guess they dropped support for XBox One at some point? I started having problems with it five years ago. I don't remember a big compatibility-break launch since I purchased it, so I'd expect it to be able to turn on + connect to their servers, or at least run the games it's already downloaded + installed.
I do own one MS game that still works: A copy of Minecraft. It took over 8 hours to figure out how to get it stop constantly asking my kids for my master MS account password. That did convince me to actually wipe all data from my Microsoft account, so I guess it was a win.
https://en.wikipedia.org/wiki/NSA_warrantless_surveillance_(...
I suspect the 2007 in the title refers to the fact that bills were passed to ban this stuff in 2007, which is when the PRISM program (also illegal domestic surveillance) got started.
(The title makes it sound like warrantless surveillance lasted from 2001-2007, but I think it means the article only covers that date range.)
reply