Today OpenAI announced that "Revenue share payments from OpenAI to Microsoft continue through 2030, independent of OpenAI’s technology progress" That "independent of OpenAI’s technology progress" fragment appears to mean that the weird AGI clause is now deceased
@simonw
-
Interesting Sandbox Product for AI Development
By
–
I would categorize your product as a "sandbox" – looks like a really interesting one
-
GitHub and Litestream: Data Backup Solutions for Developers
By
–
I have a bunch – GitHub for code and Litestream for SQLite are two, I even have a system where my PostgreSQL data for my blog is backed up as JSON data in GitHub to track history over time https://
github.com/simonw/simonwi
llisonblog-backup/commits/main/
… -
Protect Data with Append-Only Logs and Operation Replay
By
–
You can also protect against this kind of situation through data design – you can have an append-only (backed up) log of key operations that transform data, such that you can review that log and replay it should the need arise
-
Backup Security: Protecting Data Beyond Single Vendor Risk
By
–
Having your backups on the same vendor as your production database puts you at risk of your vendor banning your account for some reason (like billing messups, or detected CSAM in user-uploaded content) – having your backups held outside that vendor protects against that
-
Data Leaks from Software Bugs Cannot Be Reversed
By
–
That's true: a bug that leaks your data can't be wound back using backups
-
AI Agent Incidents: Backup Recovery Strategy for Production Systems
By
–
Backups fix that too. If the agent spent six hours rewriting your production logic good backups will let you roll back those six hours, of both data changes AND code changes
-
AI Decision Explanations Are Always Post-Hoc Guesses
By
–
Answers to why they did something are ALWAYS guesses – the original calculations have been discarded by the time you ask the question, all they can do is guess a plausible reason
-
Sandbox limitations and AI system safety constraints discussion
By
–
I want my sandboxes to limit what they can do when they are running It sounds like you and I may have different definitions of "sandbox"
-
Agents Make Surprising Mistakes Unlike Humans
By
–
Right – agents aren't humans, and the mistakes they make can be deeply surprising if you expect them to work like humans do