Where do you see the pro china side getting more and more support? As far as I can tell it's sharply swung towards maintaining independence in the past decade or two with single digit support of unification with the mainland.

https://esc.nccu.edu.tw/PageDoc/Detail?fid=7801&id=6963

https://esc.nccu.edu.tw/upload/44/doc/6963/Tondu202512.png

https://jamesoclaire.com - with a focus on mobile advertising and adtech

https://appgoblin.info -AppGoblin: mobile SDK and aso marketing, mostly self hosted

A AppGoblin report on the relative increase and decrease in mobile app SDKs from the top 50k apps in the app stores. Most of the data is from Android with some iOS. The apps are limited to those that were released before June 2025.

Check the developement tools for open source / development libraries.

Some interesting take aways:

Rive saw big increase in apps. Developer focused analytics like Datasig, PostHog saw big increases MongoDB decreased

My friend works in that period of Ottoman archives. Do you have a source or something I can share?

  Location: US/TW
  Remote: Open to Remote, In-person, or Hybrid
  Technologies: Linux, PostgreSQL, Python, mobile advertising 
  Résumé/CV: jamesoclaire.com/portfolio
  Email: jamesoclaire@gmail.com

15 years experience in mobile advertising, mobile ad buying and running teams. Excellent at hacking together new prototypes and building proof of concept software.

My main skills are around databases, data pipelines and mobile ad system such as creative automation, ad buying algorithms and mobile ad attribution. I (actually) really enjoy mobile advertising, and have built companies and projects in the space. I'm also passionate about offgrid solar, self hosting, forestry and mobile security.

How many children survived covid thanks to vaccines? How many elderly avoided getting covid due to vaccinated children.

You're right it's a red herring.

It's at least ten out of 96 deaths reviewed. If those are a random sample, then these findings are incredibly alarming.

They are not alarming without some significant context. At the very minimum, what was the health status of the people who died? People at high risk of heart issues would be more likely affected. But also COVID itself causes heart inflammation, especially in young people, so the interesting ratio is really the chance of death from Covid vs chance of death from vaccine.

10 out of 96 can be either alarming or an amazing positive outcome. Without more context, it's not actionable and mostly irrelevant.

  Location: US/TW
  Remote: Open to Remote, In-person, or Hybrid
  Technologies: Linux, PostgreSQL, Python, mobile advertising 
  Résumé/CV: jamesoclaire.com/portfolio
  Email: jamesoclaire@gmail.com

15 years experience in mobile advertising, mobile ad buying and running teams. Excellent at hacking together new prototypes and building proof of concept software.

My main skills are around databases, data pipelines and mobile ad system such as creative automation, ad buying algorithms and mobile ad attribution. I (actually) really enjoy mobile advertising, and have built companies and projects in the space. I'm also passionate about offgrid solar, self hosting, forestry and mobile security.

Here's some of the biggest mobile apps using mixpanel:

https://appgoblin.info/companies/mixpanel.com

I misunderstood this at first, it looks like the SDK & code is open source but all the billing data is sent to Flowglad servers so you don't 'own' your data?

Yeah exactly, it's a very misleading title in every dimension.

The LLMs rate of improvement has really slowed down. This looks like a minor improvement in terms of accuracy and big gains from efficiency.

14 months ago we had GPT-4 and now we have models that can get a gold medal at the IMO.

But sure, if you curve fit to the last 3 months you could say things are slowing down, but that's hyper fixating on a very small amount of information.

Yes, that is what I'm saying, that 14 months ago the rate of change was noticeably faster. Lately the new models are much less groundbreaking and increasing in the volume of output and decreasing in cost.

The private model that got gold at IMO was 4 months ago. 14 months ago we had o1-preview, we didn't have that gold medal winning approach yet. You could only say that things have slowed down since 4 months ago, but in my view that's reading the tea leaves too much. It's just not enough time and too little visibility into the private research.

it could be results of corps focusing resources on IMO in PR wars, and results is not as generalizable outside this niche.