Amazon will make an estimated $100 billion investment in capital investments in 2025, with a “vast majority” of that going to artificial intelligence. Alphabet has indicated it will invest about $75 billion; Microsoft plans to spend $80 billion while Meta intends to invest $60 billion to $65 billion in 2025.
Combined, the four companies have reported capex of almost $250 billion in 2024 and forecast this to rise beyond $300 billion this year. All of which will make some observers--who are uncertain about the payback--quite nervous.
Indeed, as Jassy noted on Amazon’s most-recent earnings call,”there aren't that many generative AI applications at large scale yet.”
The gamble is all a matter of perspective. Skeptics tend to want near-term proof of profit potential. Optimists see a generational opportunity. The former will want to see near-term financial results; the latter believe the opportunity for disrupting computing and business models is so profound they cannot allow the opportunity to be forfeited by inaction or inadequate vigor.
“AI represents, for sure, the biggest opportunity since cloud and probably the biggest technology shift and opportunity in business since the Internet,” said Amazon CEO Andy Jassy. Who siad AI is a “once-in-a-lifetime” type of opportunity.
Sundar Pichai, Alphabet CEO, said the AI opportunity was “as big as it comes.”
Still, the arrival of DeepSeek does offer the promise of lower model costs, Jassy indicated. “I think like many others, we were impressed with what DeepSeek has done.”
And Amazon seemed notably “impressed with some of the training techniques, primarily in flipping the sequencing of reinforcement training, reinforcement learning being earlier and without the human in the loop.”
As typically is the case for all of computing, allowing machines to work without the constraints of human processing speed tends to speed up progress.
Jassy also addressed the fear some have that lower model costs (inference and training) will lead to smaller markets for AI products and infrastructure.
“Sometimes people make the assumptions that if you're able to decrease the cost of any type of technology component, in this case, we're really talking about inference, that somehow it's going to lead to less total spend in technology,” Jassy noted. “And we just, we have never seen that to be the case.”
Indeed, computing economics suggest that vastly-lower prices per compute cycle or storage costs lead to people and businesses creating new use cases that formerly were cost prohibitive, with the result that overall technology spend grows over time, Jassy argued.
Customers “get excited about what else they could build that they always thought was cost-prohibitive before and they usually end up spending a lot more in total on technology once you make the per unit cost less,” he added.
“I think that is very much what's going to happen here in AI, which is the cost of inference will substantially come down,” Jassy said.
So, as often is the case, financial analysts and firm leaders are asking different questions. Financial analysts always are focused on the next quarter. Firm leaders necessarily are focused on the longer term, especially when a disruptive new industry possibly is being born.
It’s an old story. Entrepreneurs, investors and product developers must accept risk as the price of winning. Lawyers, accountants and financial analysts all want to “de-risk” as much as possible.
And we are some ways from proving the visionaries or the risk managers correct.
If the former are correct, then the new equivalents of market leaders in search, social media, e-commerce and entertainment will emerge. And those potential leaders will have to invest heavily now.
If the latter prove correct we will see possibly-massive investment losses and bankruptcies for many contestants who overspent too early, even if the market does eventually emerge as expected.
On the other hand, in the computing field we have often seen legacy leaders displaced because they failed to invest in the emerging substitute technologies.
No comments:
Post a Comment