Thanks for this write up. It’s finally the time of the generalist!
I have also been blown away by Claude code. I built an entire stakeholder CRM in literally 20 minutes the other day, making it easy for me to create and send project updates to different stakeholders, something which can take hours of my time (similar to handover exercises!)
Q: How do you deal with the sheer amount of content created in Gen AI work? Something I’ve noticed with junior team members, the ability to create outcomes quickly mean 10x previous outcomes are indeed created - not all of which are needed or good. E.g Do we need a report where a memo does the job? I know this is where taste and discernment come in, but how do you handle this issue whilst we develop it. Is it a case of assuming most reports will be read by AI anyway?
Similarly, like the endless online swiping of online dating I think there can be a tendency to keep searching for the next better version of what you’ve created, now it’s so easy. Do you think being AI native means really committing to the old adage ‘done is better than perfect’? Or will we need a new mantra for the AI age?
This is really interesting framing. I’m very much a done is better than perfect person, and so for me the issue is more getting to a workable version then not doing the refinement. I’m interested in what it would be like pairing with someone on our team who is a lot more of a details finisher, and if I do v0 of the build then someone else takes it to v1.
Re the wastage / lots of outcomes side. I think taste is the key here. What we don’t want to get to is just volume slop, there needs to be filters. I think fortunately also clients are a good buffer in that process too.
Love the sound of your CRM. Is it something you trigger yourself or do you have the whole process automated now?
Thanks for this write up. It’s finally the time of the generalist!
I have also been blown away by Claude code. I built an entire stakeholder CRM in literally 20 minutes the other day, making it easy for me to create and send project updates to different stakeholders, something which can take hours of my time (similar to handover exercises!)
Q: How do you deal with the sheer amount of content created in Gen AI work? Something I’ve noticed with junior team members, the ability to create outcomes quickly mean 10x previous outcomes are indeed created - not all of which are needed or good. E.g Do we need a report where a memo does the job? I know this is where taste and discernment come in, but how do you handle this issue whilst we develop it. Is it a case of assuming most reports will be read by AI anyway?
Similarly, like the endless online swiping of online dating I think there can be a tendency to keep searching for the next better version of what you’ve created, now it’s so easy. Do you think being AI native means really committing to the old adage ‘done is better than perfect’? Or will we need a new mantra for the AI age?
This is really interesting framing. I’m very much a done is better than perfect person, and so for me the issue is more getting to a workable version then not doing the refinement. I’m interested in what it would be like pairing with someone on our team who is a lot more of a details finisher, and if I do v0 of the build then someone else takes it to v1.
Re the wastage / lots of outcomes side. I think taste is the key here. What we don’t want to get to is just volume slop, there needs to be filters. I think fortunately also clients are a good buffer in that process too.
Love the sound of your CRM. Is it something you trigger yourself or do you have the whole process automated now?
Yes, effect of AI is overwhelming. Please go through this article to know how AI aided in scaling a tedious task.
https://substack.com/@uxlivinglab/p-193230601