Artificial intelligence (AI) has been the cat’s meow in terms of big-dollar investment over the past couple years, but the enormity of these buy-ins has some investors wondering when—or if—the payoff will come (1/13)

Investors have plowed billions into AI large language models (LLMs) and related projects. There’s endless hype regarding the technology’s abilities and lofty projections about the impact it will have on the global economy and on society (2/13)

But a recent Goldman Sachs report (‘Too much spend, too little benefit’) called these projections into question, with some analysts suggesting AI’s impact over the next decade won’t be anywhere near as great as currently imagined (3/13)

https://web.archive.org/web/20240629140307/http://goldmansachs.com/intelligence/pages/gs-research/gen-ai-too-much-spend-too-little-benefit/report.pdf

While most analysts believe that AI will ultimately prove a game-changer in a number of fields, including scientific discovery, research & development and new product testing (to name a few), some believe the shorter-term horizon will be far less dramatic (4/13)

Current AI architecture can increase efficiency through automation, but you eventually hit a productivity ceiling—at a staggering cost—and truly transformative changes could prove beyond the capabilities of existing models. And they NEED to fix the hallucination issue (5/13)

Meanwhile, the costs of developing and implementing AI are staggering and might not significantly decrease over time—and could soar if China’s bellicose stance re Taiwan (and its chip sector) moves beyond rhetoric to action—making further AI inroads cost-prohibitive (6/13)

Not all analysts are as fiscally pessimistic, arguing that AI’s current cost/benefit ratio isn’t that dissimilar from previous tech investment cycles, that technological breakthroughs are always around the corner and AI’s ‘killer app’ will eventually emerge (7/13)

None of the ‘Mag 7’ companies investing heavily in AI are struggling, at least, not according to their Q2 profits. But there’s definitely an arms race here, a fear of being left behind, so the spending spree will continue whether or not AI revenue keeps pace (8/13)

A possible tripwire is AI data centers’ insatiable energy appetite. Without some tangible proof that this technology will be as game-changing as its hype, will governments approve the massive infrastructure development needed to ensure AI continues to grow? (9/13)

AI’s energy requirements could prove even harder to justify down the road as the planet continues to heat and wildfires and other natural disasters increase in both frequency and intensity. A more efficient way to burn the world isn’t what’s needed here (10/13)

Blockchain technology has been subjected to similar criticism due to Satoshi launching Bitcoin 15 years ago and yet, as far as the wider public knows, there’s been no blockchain ‘killer app,’ just endless financial speculation, rug pulls and ‘number go sideways’ (11/13)

I have faith that AI will live up to the hype, just as I believe the public, governments and enterprises will eventually see the benefits that an endlessly scalable, utility-focused blockchain has to offer (12/13)

https://aerospike.com/news/bsv-association-and-aerospike-achieve-100-billion-bsv-blockchain-transactions-per-day-on-teranode/

Even better, AI will combine with enterprise blockchain and IPv6 to enable Metanet, the more inclusive and dynamic internet that enables cost-effective and intermediary-free instant micropayments between creators and consumers. Can’t get more ‘killer’ than that (13/13)

Calvin Ayre
Aug 14, 2024
https://x.com/CalvinAyre/status/1823683577231397015?t=CMPqX1iQ00zflkckdYpd0Q&s=19