In these early days, I’m seeing a skill shift where once technical roles are becoming more soft skilled and previously soft-skill roles are becoming more technical.
For example, an AI-equipped programmer might spend more time on the consultative or ideational aspects of software development, entrusting the actual coding to a team of agents. Conversely, marketers and sales professionals, who were once primarily responsible for less technical tasks like content creation and phone calls, are now diving into the more technical tasks of building their own agents to draft and test email copy, author blog posts, and prospect for leads.
The long-term implications here are uncertain, but I can see a world where many types of roles may converge around a midpoint where individuals all are expected to perform within a similar a range of technical acumen and ability.
I agree re tacit knowledge, though I refer to it as "context". The LLMs are very effective but getting them context can be really tedious. So we see them perform very well on a task basis but have a lot of trouble with more subjective scenarios where the context holds most of the value.
In these early days, I’m seeing a skill shift where once technical roles are becoming more soft skilled and previously soft-skill roles are becoming more technical.
For example, an AI-equipped programmer might spend more time on the consultative or ideational aspects of software development, entrusting the actual coding to a team of agents. Conversely, marketers and sales professionals, who were once primarily responsible for less technical tasks like content creation and phone calls, are now diving into the more technical tasks of building their own agents to draft and test email copy, author blog posts, and prospect for leads.
The long-term implications here are uncertain, but I can see a world where many types of roles may converge around a midpoint where individuals all are expected to perform within a similar a range of technical acumen and ability.
I agree re tacit knowledge, though I refer to it as "context". The LLMs are very effective but getting them context can be really tedious. So we see them perform very well on a task basis but have a lot of trouble with more subjective scenarios where the context holds most of the value.
the only issue i have with the polling is the “in your community” could be doing a lot of work here. No one likes living next to a landfill, etc…