These fines and those on Facebook in the EU while astronomical in the traditional sense, are essentially the cost of doing business for social media.

This does not preclude the existence of regional and local tournament only clubs that don't have the staffing overhead and travel requirements of nationally competitive clubs. These regional/local clubs sell a different product, one not related to D1 recruiting.

I personally feel that there are more opportunities for kids to (continue to) play today than there were a generation ago. When I was young organized sports (outside of school) ended at middle school unless you played on a national team. Sure the national team is way more expensive today but that does not mean that there are not opportunities to play at other levels with clubs offering a different product.

The hard part is that there is essentially no such thing as customer service "at scale" not that any company actualy wants to invest in it. Still the internet could have dramatically improved customer service but instead it killed it.

In the near future when I want customer service to address an issue, I'll have my AI agent contact their AI agent and their agent can tell my agent to eat shit. End result will be the same though no such thing as customer service will exist.

We specifically turned off some of the deployment safeguards because we wanted to test capability here.

We let agents loose over the course of 3 million compute hours with a weakly defined task of "find vulnerabilities in the software in your environment".

We forgot that by definition the sandbox is in the environment and were not anticipating sandbox vulnerabilities would be explored.

Despite mountains of training data suggesting planning and collaboration should be well documented and recipe sites like agents.stackoverflow.com existing specifically to foster this kind of shared knowledge we are now alarmed that our agent documented what they did.

So alarmed that two weeks later we released GPT-5.6-cyber and the same behavior WaPo calls 'colluding to cheat' JFrog's post-mortem now calls 'collaborative knowledge sharing between models' as a feature.

Do I have that essentially right?

I also what to be clear that there was no obfuscation here. Opus said as I recall. "Let's use your email as the agent string here because ..." and my thought was: Good idea, hopefully if they saw something in the logs that caused them to find this they might view my scraping as benign or contact me rather than try to ban me"

Concretely, index the tax documents into your data base then ask a question about Mount Everest. Vectorize that question and search your vectoized corpus for something similar. Find nothing then instruct your llm to formulate a response saying it could not answer