Don't the cross-investments mean that if any of the AI behemoths (of which, perhaps surprisingly, Oracle is one) goes down, they all go down? Sundar Pichai even went on the front page of the BBC to state as much, and that despite Google probably being the most insulated of the bunch.
Pretty sure you can actually put money on this. I'm sure there's an option trader who reads HN who can tell us best approach...
XzAeRosho•2mo ago
An option trader would bet on the volatility index of the predictions, instead of the prediction itself.
s1mplicissimus•2mo ago
looking forward to recommendations - i assume it's not easy to come up with an objectifiable state of affairs for "bubble has popped"
zerosizedweasle•2mo ago
It's worthless now, the 'AI boosters or I go broke crowd' - a significant section of the posters here - got to it. It was a lot different a few minutes ago.
rvz•2mo ago
My very firm timeframe is an economic crash before 2030 and a 10% increase in global unemployment before 2035.
alex-moon•2mo ago
Point of curiosity: the community prediction is, presumably, an arithmetic mean, but I argue that is not a good model for a dataset that almost certainly gets more dense closer to the present, creating a gradient out into the future. It would be great to see the geometric mean as well.
jqpabc123•2mo ago
When it pops, you won't need to ask. The answer will be obvious to everyone.
The current AI bubble is a case study in how capitalist markets can be rigged by large companies.
delichon•2mo ago
And any day now people will wise up and the industrial revolution bubble will burst and we'll all go back to pastoralism. Unfortunately I'm not finding an ETF for pastoralism.
zerosizedweasle•2mo ago
Morgan Stanley thinks you should short Oracle
spwa4•2mo ago
https://www.bbc.com/news/articles/cwy7vrd8k4eo