After recovering from the surprise, I remembered that I had a brief statement about expertise and appropriate levels of detail in my user instructions.
@patio11
-
ChatGPT provides surprising exercise advice recommendations
By
–
Asking ChatGPT for exercise advice and, after the usual fairly anodyne bulleted list of not-too-much-alpha, got something which surprised me:
-
Real Estate Analytics: From Manual Studies to SQL-Powered Insights
By
–
You wonder what anybody with just a bit of verve plus any source of good trip data could do in e.g. real estate. A while ago traffic studies were $$$ and their results archived for decades and poured over manually. Now you can get a map arbitrarily colored at the speed of SQL.
-
FAA conservative approach to novel aviation technology regulation
By
–
FAA is an extremely conservative regulator and generally takes the position that anything novel which cannot be proven to not interfere with flight systems is presumptively a threat to flight systems. This exists in some tension with other goals we have for flight infrastructure
-
Cost efficiency in AI: scalability as business imperative
By
–
Overwhelmingly “Eat it as a cost of doing business at most relevant margins; unless it is scaled it isn’t worth the cost of human attention. However, if it is scaled it is another thing entirely.”
-
Balancing Fraud Prevention with User Transparency and Enforcement
By
–
This involves solving for competing interests, attempting to increase the legibility of the systems to users, minimizing cost of enforcement actions, dealing with the reality that fraud is a cause of many but not all disputes and some people are as honest as their options, etc.
-
Payment Rails: Dispute Arbitration as Competing Virtual Jurisdictions
By
–
An underappreciated thing about payments rails is that one dimension they compete on is dispute management. This often involves arbitration systems which are, essentially, privately administered legal systems. And those virtual jurisdictions compete like real ones do!
-
Delivery Services Deploy High-Frequency AI Dispute Management
By
–
Less impressed about not getting the coffee for but I continue to be impressed that the delivery services now have High Frequency Dispute Management and that if you say “Didn’t get a coffee” then you get credited back the money in less than 30 seconds with no human involvement.
-
Insurance Regulation Fragmentation Prevents Consumer Decision Making
By
–
Insurance regulation is byzantine levels of sharded, and while in principle one could aggregate the information, customers would probably not use it to make decisions at the retail level and so nobody puts in that work for that purpose.
-
Expert Data Analysis vs Personal Estimation in Decision Making
By
–
Who are you going to believe, your best estimate or a team of statisticians who have 20+ years of full claims data for their firm plus > 1000X that (but hazier) in the broader industry and do nothing but study variations on this problem? And they *don’t even have to be right.*