Ozan D.

When it comes to working with agents that are more empowered to take actions on your behalf (like Claude CoWork for instance) I find the "responsibility ladder approach" from leadership/management best practices highly useful.

Best to think of AI as a thinking partner rather than as a tool and approach it with a very high level of critical thinking and scepticism. Which is what most people find very difficult to do. In our tuning behind the scene with AISA (AI skill assessment) our model, for example, looks at signals to understand whether the person is seeing AI as a one direction tool (like a hammer) or sees it as an "agent" or a "partner". This perspective alone has been incredibly helpful in telling apart people that are okay vs great at using ai.

Hey Anith, I wonder - whether you are talking about the amazing cutting edge of the tech you've built enough in these reach outs or does it come across to them as just another spam?

I broke up with ChatGPT and would never go back! Opus 4.6 is the only model I trust "enough" to work with.

I'm building something new and it's in my to do list, but I'm working to build some more hype around it before I go on there.

I had done it with a couple of products in the past and got both good and bad results. The key bit is to build hype around it and make sure you are posting there at the point where you have some initial social proof!

The answer is - as is the case with most things AI- Partly. What we are most likely to see is a fracturing of all role descriptions and job functions.