-0.008592
Relative Brier Score
21
Forecasts
5
Upvotes
Forecasting Calendar
Past Week | Past Month | Past Year | This Season | All Time | |
---|---|---|---|---|---|
Forecasts | 0 | 0 | 21 | 21 | 21 |
Comments | 0 | 0 | 5 | 5 | 5 |
Questions Forecasted | 0 | 0 | 14 | 14 | 14 |
Upvotes on Comments By This User | 0 | 0 | 5 | 5 | 5 |
Definitions |
Reducing probability - it looks like Iran is currently trying to walk a fine line between maintaining its international credibility while not provoking Israel too much that they declare war.
Indicators are pointing towards an escalation between Israel and Iran short of outright war. I would not be surprised to see this being escalated to a formal declaration of war in the next year or so.
As previously described, Microsoft is trying to hedge their bets by trying to retain access to Chinese talent and not pausing recruiting while also setting up the option to relocate to Vancouver if it becomes politically untenable to maintain a presence in China. A lot would hinge on the upcoming presidential election and the messaging around China - I wouldn't be surprised to see political pressure from both sides for this to happen in order to appear tough on China. Perhaps with the inevitable Trump win we would also see increased pressure to decouple from China so this could very well happen in 2025/6, though it is not entirely inconceivable that Microsoft will see the writing on the wall and proactively pull out sooner if sufficient political pressure is applied.
Confirming previous forecast - it seems that FB is looking to detect and label images with AI-generated watermark but nothing on detecting non-watermarked images/text - fundamentally the issue is that AI generated text has reached the point where it is essentially indistinguishable from normal human language.
Why do you think you're right?
The main consideration is that Huawei has explicitly denied this [1] last month at MWC24.
Other things to consider is that the desire to disaggregate parts of your networking stack is mainly a political one - why have a single black box running your comms where you can have multiple black boxes provided by different people doing the same? Jokes aside, disaggregating your network stack creates extra latency and cost - at the end of the day a decentralised mesh network (which is what it seems to be) is always going to be more complicated to implement than a simple hub and spoke network.
There's also the other consideration that Huawei is betting on seeing this fail - ORAN has been widely touted as this decentralised alternative to Huawei and it would be in their financial interest to see a return to the status quo where they have built up an existing moat.
Additionally, it appears that they are prioritising their growing but still constrained advanced node capacity towards building AI accelerators with networking taking a back seat to this. Also, the Chinese chip industry as a whole is currently extremely skittish about releasing any new domestically manufactured advanced node products to the international market for fear of triggering further retaliatory sanctions. Due to this, Huawei has mainly been prioritising the domestic networking market and I don't see them being especially keen to drum up further international interest for their networking products by engaging in this programme (especially if you consider that their brand is already politically toxic to a majority of members of the consortium - I don't see how participating in this can help redeem them)
[1] https://www.telecomtv.com/content/open-ran/huawei-denies-open-ran-support-speculation-49795/
Why might you be wrong?
If the Chinese mobile operators decide that Open RAN can result in a better overall product than Huawei and therefore decide to jump ship - this might motivate Huawei to change course.