@simonw @solidstate_birb In last week’s issue of Synthetic Work, I mentioned exactly this happening to me: having access to both Copilot and GPT-4, and in need of some Javascript code (I’m not a dev), I found myself asking the question to the latter rather than the former.

@simonw @solidstate_birb As I wrote in the newsletter, I believe there’s a silent, ongoing fight to establish what is the job-to-be-done in the mind of the users that are exposed X number of LLMs.

@emollick 100%. Prompt massaging underneath pretty UIs is not a long-term competitive advantage.

@emollick As a former industry analyst (Gartner, I left almost 10 years ago), I can confirm that product categorization has an enormous impact on the success of a company.

Self-assigned vs “inflicted by analysis firms” has radically different pros and cons nobody ever talks about.

Dusty playbooks get recovered from the archive:

https://t.co/8i36nyA54s

@danielgross https://t.co/G0LCEaf4TY

At a dinner last night, I met the owner of a small creative agency here in London. He tells me he uses GPT-4 every day to speed up the delivery of various collaterals. Coming from a disadvantaged family, he sees AI as a great equalizer for people that can’t afford top education.

Oh, this looks fantastic. On my must-try list for the weekend. https://t.co/zYBTazUJMS

It was bound to happen. I have first-hand experience fine-tuning Salesforce CodeGen and I know how good it can be. @jamesravey did a great job here. https://t.co/WEp3aBK75u

It will be interesting to see how the “MidJourney dilemma” applies here and how it will be addressed >

China slaps security reviews on AI products as Alibaba unveils ChatGPT challenger https://t.co/8xUZsNR7lx