AI and Bitcoin
The two things boiling the planet for nothing more than powering a hype machine. make me ashamed to be associated with IT (And yet tend to be the two topics people always bring up with anyone in IT)
The AI datacenter spending splurge looks likely to continue in 2025, with Microsoft alone saying it will invest $80 billion this year on building out infrastructure to train and deploy AI models. In a blog post that reads almost like a call to arms, Microsoft president and vice chair Brad Smith described AI as a "golden …
And if you ignore AI now or keep pointing out how shit/danagerous it actually is, you get ignored at work because they like pissing money on it and riding the hype train but fuck whits in charge with no IT knowledge. Mainly because one director (who shouldn't be in that role) is using CoPilot to write all his reports for him.
"We have the opportunity as a country to equip all Americans* with the skills needed to use AI to pursue higher-paying jobs and more successful careers" = please daddy gubbmint give me tax dollars.
*subject to terms and conditions - exceptions include anyone with a net worth less than $25 million USD, uses non-conforming pronouns, is disabled, etc.
The money Microsoft has invested feels like it's getting towards "bet the farm" and "too big to fail" level.
If the AI bubble pops, I fully expect Microsoft to find themselves in serious financial problems very quickly. The resulting court cases are going to be legendary.
Still, IBM could always dust off OS/2 for all the management people who react to the words "open source" the same way arachnophobes do to a picture of a tarantula. So there's that, I suppose.
"Even if there really is a market for this kind of 'AI', it's nowhere near large enough to ever pay back this kind of money."
You might think so, but I'm not so sure based on some rough numbers:
a) MS revenues for 2024 near enough $250bn.
b) $80bn investment needs say a return of 12%...
c) requires an extra $10bn a year of sales (assuming next to nil operating cost),
d) To wash its face and make a profit, MS have to persuade their customers to pay them 4% more.
Last year MS revenues were up by just shy of 16%, year before 8%, year before 18%, so I'm seeing 4% for the dubious benefits of AI as being easily realised.
I think your error is in assuming that there's any link between the spurious benefits AI might bring, and the amount credulous corporate fools can be persuaded to pay.
A tax write-off AND the prelude for 50% raises for SatNad and his BOD.
Which will in turn lead to price rises for everyone else. That's capitalism in 2025 with Trump in charge. Tax rises for us, new jets for them. Trump 2.0 Yay!
This says everything.
https://www.youtube.com/watch?v=fAuI1lvSBqc&pp=ygUJZG9sbGVtb3Jl
How long after the 20th will he be allowed to carry on doing that is anyones guess.
Investment of this scale in AI is beyond most countries. How can the UK (or any other country than the US or China) - anything smaller than FAANG etc - compete with this level of investment (and its potential yield)? It seems like we're beyond the tipping point of potential competition. Non-FAANGs will need to focus on efficiency rather than scale?
Even if there really is a market for this kind of 'AI', it's nowhere near large enough to ever pay back this kind of money.
Of course there is. If we assume AI only replaces call center jobs, and it clearly can start to do that now, then the spend discussed in this article gives roughly a 31k break even cost cut per former employee. If we deprecate that over a 5 year shelf life, then anything over about 6k per year in wage costs is all cheddar. Just from call centers.
I'm not a fan of AI, I think we're going to prematurely push a lot of people out of work with it, which will predominantly affect Gen Z (just finished racking up uni debt for questionable degrees to do jobs that won't exist for long), though clearly not all of Gen Z. In the future it'll just be a productivity tool, but between now and then, lots of jobs are going to vanish in lots of fields.
Except that it's not good enough for call centres, and the break even point is more expensive than the humans.
Anyone considering this is currently using a cheap offshore callcentre, as they're clearly wanting to cut costs more than is reasonable.
The cheap offshore callcentres cost about $3k USD a year in wages per call handler. So assuming the callcentre supplier is using a 100% markup, the AI capex recovery alone is equal - nothing for running costs, let alone profit.
(And yes, these pay figures are pretty scarily low)
FecalBarf, Xshitter and TroothFarcical.
Their bots will post memes of themselves, chat with other bots and the whole internet will collapse in a heap as they alternative switch between praise and slagging each other bot off.
It will eat itself (hopefully) and the rest of social media.
Both statements can be true, to some extent at least.
Its highly likely simple administration roles, such as those found in call centres and much of the public sector back offices, will be replaceable with a suite of AI agents. Its equally likely that doing so will throw off some much better paid much higher skilled roles in training and tuning those agents, just in lower volumes.
If you're one of those people that can "complete the assigned tasks in 2 or 3 hours" then netflix the day away, then this is your comet.
You gotta speculate to accumulate, right? May the Best of the Best of AIs Win Win ..... as they surely most definitely undoubtedly will, and as IntelAIgently Designed by Advanced Virtual Machines to do.
What next to do to not crash global markets and prevent revolutions and dissuade radical pogroms and incendiary programs is the next small step for Man and Mankind requiring giant leaps from Man and Mankind into what is quite clearly evident to any and all akin to Alien Invaders, be the Great Future Unknown Unknown.
"AI offers not only new tools for people's work but also new ways to help people learn almost anything. We have the opportunity as a country to equip all Americans with the skills needed to use AI to pursue higher-paying jobs and more successful careers.
If your job is a meaningless pusher of paper, or low quality junior coder and you've learnt you can bunk off early by getting AI to churn out something that vaguely resembles what you are supposed to be doing, you are going to be in for a shock if you think you'll be paid more or have a successful career. If you can get away with this, it won't be long before the company realises and replaces you all together.