Comment by samhuk

1. Internet services become increasingly fractured. China bans more services. Then Russia joins in, with Middle-East following suit. US government gets jealous and so "Should US Ban TikTok?" becomes the New Thing that is talked about for a good part of 2023.

2. Ukraine fights on but cannot last forever against a country with a much larger population and, arguably, industry (capable of war). Tensions rise, but many governments drag feet out of fear of killing their already covid-striken economies.

3. Austerity in UK becomes the only solution as budgets with 30 billion pound black holes kind of can't exist. NHS collapse, that has been occuring for ~10-20 years, accellarates much faster. Productivity declines due to sicker population. Vicious cycle becomes apparent over next ~5 years as sicker workers -> lower productivity -> lower GDP -> worse healthcare -> sicker workers -> ... Will be blamed on brexit and/or "lazy people", as was the case ~10 years ago when "benefit fraudsters" became the new group that the UK government told everyone to hate'n'blame.

4. EV uptake misses the current jubilant expectations as the first-world eventually catches on to the absolutely dispicable atrocities occuring to mine {metal} (e.g. Cobalt, Tantalum, etc.) for {battery dependent thing}. EV companies employ FUD to divert blame. Perhaps ask government to help silence discussion, Twitter Files like?

5. Scalping (i.e. PC parts, etc.) will still be a problem (although not as bad) since there is literally every incentive for companies to not do anything about it. Many scalping services, scripts, github repos, etc., spawned from covid times, lowering the bar so Joe Bloggs can do it, spawning what will be an endless cat-and-mouse rat-race to the bottom until {something} happens.

6. Small-time logo designers and all related industry deflates as stable diffusion et al automates a good chunk of it, lowering the bar.

To end on lighter notes:

7. {new product} on Show HNs will still be dismissed by some as "simply achievable with rsync".

8. Work on nuclear fusion increases as it becomes less of a toy/pipe-dream and machine learning enables rapid progress and iteration.