5 ms·
Disagree with the first myth (productivity) because I've already implemented LLM-related automations that already have saved a lot of time. Both involved summar
by bschmidt1 2y ago
Disagree with the first myth (productivity) because I've already implemented LLM-related automations that already have saved a lot of time. Both involved summarizing a small amount of information and pre-filling a text input in a messaging app with a suggested message to send to a client. The user can still edit the text and is ultimately responsible for pressing "Send", but the LLM takes you 90% of the way there by referencing past messages and referring to completion criteria to write the message you would probably write.
Tasks like replying to emails, applying for jobs, and sending daily updates are deceivingly draining. Creative writing is in itself a task, and when the content is repetitive and just summarizes other work you can automate most of that thinking/language work. Why spend 15-20 minutes on Theusarus.com thinking of a novel way to say something - offload that to a language machine.
Would extend this to customer service in general - probably won't have cashiers, drive-thru order takers, or customer support chat as human jobs very much longer.
- cratermoon 2y agoIf someone puts little to no effort into communicating, why should I put any effort into listening to what they are trying to say?
- bschmidt1 2y agoPeople don't care if customer support is AI if the problem gets solved sooner/more easily. Similarly, some people are more receptive to AI therapists than human therapists - anyone who doesn't feel comfortable sharing details with a person who might judge them, but would still value the intelligent analysis of their situation. Not sure it logically follows that automated or low-effort messages inherently have no value. In some cases, receiving a message where little to no effort into communicating was placed may be better - for quick responses if nothing else.