In the age of AI, taste is the differentiator.
But how do you channel taste into the things you build?
@bkase_ and I discuss 4 emerging practices:
Shifting left inside agent harnesses, expert language, tool interfaces, and oracles.
01:54
@_lopopolo's (OpenAI) talk on harness engineering and shifting left
05:45 Birgitta Böckeler on guides and sensors in harness engineering
13:11
@rauchg (Vercel) on mastery of language in the age of AI
17:04
@bkase_ on borrowing expert language from an additional domain (algebra) with better abstractions
26:06 MCP, CLI tools, Bash, and code mode
32:07
@trq212 (Anthropic) on how Claude can express a workflow through code
35:47
@mattzcarey (Cloudflare) on a code-mode-style approach to search
36:37
@bkase_ on when using code mode is better
40:48
@justincormack (former CTO at Docker) on building an S3 clone using an oracle at
@tessl_io's AI Native DevCon London
Would love to hear what practices others are using to channel taste into what they build.