| user: | andyk |
| created: | August 17, 2012 |
| karma: | 257 |
| 1. | |
| 2. | #1 agent on swe-bench wrote 7% of its own code(aider.chat) |
| 3. | |
| 4. | AI benchmarks should be like unit tests(andykonwinski.com) |
| 5. | OpenAI Dropped Work on New 'Arrakis' AI Model in Rare Setback(theinformation.com) |