olavo_sg

Putin is running short of resources and must think very well if he wants more enemies.

Kazakhstan 
It seems like the leader is relatively close to Putin and both countries are profiting. I dont think Putin would change it right now.  
https://www.chathamhouse.org/2024/02/russias-influence-kazakhstan-increasing-despite-war-ukraine

Georgia
A little bit higher odds, but there is a Pro-Russia administration right now, which may reduces Russia's incentives to intervene.

Armenia
Lately is getting closer to the West, which may prompt Russia to intervene. 

Moldova
Also getting closer to the West, pro european party won the last poll. PM is pro europe. https://www.reuters.com/world/europe/moldova-appoints-pro-eu-munteanu-prime-minister-2025-10-31/

confirmed previous forecast

- energy exports wont go down soon; it dominates total export values, leaving intra-regional trade as a relatively small share.
- limited economic complementarity among regional economies. Many countries in the region export similar commodities  and import manufactured goods, which reduces the natural incentive for intra-regional exchange.
- lack of infrastructure and connectivity
- geopolitical rivalries (algeria vs morocco for instance) and related or not border restrictions, tariffs, divergent regulations,

Confirmed previous forecast

Passage of time  + no signals of such an event

It is unlikely that the O-RAN Alliance’s specifications will be fully integrated into 3GPP standards before January 2027. Although both organizations have maintained cooperation and alignment, there is no indication that 3GPP intends to formally incorporate O-RAN’s technical specifications into its own releases within that timeframe. The standardization process is complex and slow, and 3GPP’s current focus on 5G evolution and the early phases of 6G makes full integration by 2027 improbable.

A partial or limited adoption, however, is more plausible. Some O-RAN concepts or interfaces may be referenced or aligned with 3GPP standards to improve compatibility, particularly where interoperability benefits network deployment. It is also likely that O-RAN itself—or a similar organization—will continue to play an active role in promoting open interfaces and ensuring that different vendors’ equipment can operate together. This focus on interoperability aligns with broader industry goals of fostering integration, innovation, and greater competition within the mobile network ecosystem.

I recommend reading this article.

Expanding the training dataset is one of the possible strategies to improve AI models, although not the only one. In this context, given the fierce competition at the frontier of AI, it’s likely that some players will invest heavily in scaling up their models.

However, there’s a barrier related to computational power. Colossus, xAI’s titanic data center in Memphis, is currently considered the only one capable of pushing through this frontier.

Given the time frame of the question, I find it quite likely that one of the major U.S. players will reach 10^27 FLOPs. I also see a high chance of this happening among Chinese players such as ByteDance and Alibaba.