No, Microsoft Does Not Invest $80B in Hey Hi (AI), It's Just Full of BS, Lies, and Over $80B in Debt
They want more "defense" contracts from Donald Trump (like in the first term), i.e. de facto bailout based on the false narrative of "arms race"
Fictional numbers about "worth" and "investment" are a subject we wrote quite a lot about in 2024. Both "Open"AI and Microsoft kept floating totally fictional sums of money, referring to some distant future dates (years) when nobody will still remember the false promises. It's just the art of storytelling to persuade investors to keep money in the hat, which will eventually catch on fire.
The boss of Nadella (Microsoft president Brad Smith) did the same storytelling a few days ago and a lot of "media" (or LLM slop operations) did the casual parroting, e.g. [ 1, 2]. The latter says "Microsoft expects to spend $80 bn on AI this fiscal year", but the BS artist is making up numbers to distract from Microsoft's debt and massive losses in this area of vapourware. It's also a massive legal liability - a subject that has apparently left this whistleblower dead (his mother says he was murdered).
Expect many more of those fake numbers and fake promises. Don't repeat them, ridicule them. A year ago they talked of a "trillion" in investment or over 100 billion per year. Nobody even remembers those false promises anymore!!
That's just about as laughable as the "hey hi" (AI, LLMs aren't even AI!) "era" or "revolution" or "arms race" narrative. This merits ridicule, not adoration. There used to be gutsy media that would eagerly call out Microsoft on its BS. Where's this media now? Speaking of this hilarious "arms race" narrative (as if wars are robotics and robots are "hey hi" or some nonsense like that), psydruid has just told us in IRC that Microsoft is "playing on nationalist sentiments to camouflage that it's failing all over the place, just like Linux is essentially turned into an American project."
Well, that's in Microsoft's lobbying blog (blogs.microsoft.com/on-the-issues/2025/01/03/the-golden-opportunity-for-american-ai/), in case someone wishes to copy and paste the address.
And "even with a lobbying blog, that you don't need even mention," psydruid has responded, "you can point out that Microsoft is playing on US nationalist sentiments, making its software and decisions unsuitable for the rest of the world and increasingly dangerous to its future [...] the mask has fallen off also at the Microsoft-infested Linux Foundation" (at the Linux Foundation Microsoft nowadays exercises the most influence, which it basically bought). Also many others... OSI and Mozilla, to name a couple.
Mozilla is already turning Firefox into slop-ware (LLM slop software) and there's this good response to the introduction of Mozilla’s Orbit: "As a reminder, genAI tools don’t summarise, they shorten. If you don’t understand the difference, you’ve probably been spending too much time talking with ELIZA, or autocomplete on your phone, or that stochastic parrot your neighbour bought who keeps flying in and leaving feathers everywhere because you left your window open for fresh air and haven’t got around to installing one of those flyscreens yet. Parrot screens?"
This whole "genAI" thing is a passing fad or hype because many of these tools are neither new nor innovative. They just managed to con some people into endless money-pouring, based on fictional narratives of impossible economics, based on infeasible capabilities pushed forth by intentionally false marketing. █