"These agents excel at automating repetitive tasks, allowing human workers to focus on strategic activities that require creativity and critical thinking."
Curious, do you actually believe this? I see this exact phrasing used by so many AI proponents and always wonder what it means because not one person has been able to explain what knowledge works look like in the absence of the so-called repetative tasks. The thing is, in most white-collar jobs, strategic thinking and creativity are an occasional requirement, not the norm. I'm not saying this is okay, but it means AI agents, if they are as effective as purported, will definitely gouge a big hole in corporate America and the bleeding will be hard to stop.
Fair question, Jim. I’ll defer to Lakshmi as primary author to respond in depth.
My thought is that this statement is generally true of all automation — which has its limits. An agent that incorporates AI can potentially stretch the limits of what non-AI automation agents can do. It will still have its limits, not just in what it can do, but what it should be *allowed* to do unsupervised and without recourse.
And the humans who design automations that use AI agents need to be even more sensitive to what those limits are, and what new or worsened risks the use of AI introduces, and they must design and test and monitor accordingly. There are still ‘no silver bullets’.
I agree with you that it would be helpful to pick specific white-collar jobs, break down the everyday tasks, look at where automation might be useful, and assess whether AI is needed & what it contributes & where it falls short. We may find that there are nuances to a role that an AI agent’s training will not handle well. My own recent experiences with customer support bots bear this out.
One of my favorite quotes: “In theory, there is no difference between theory and practice. In practice, there is.”
Right now it seems too much like solutions in search of a problem. I've worked in tech for 20 years and have automated many tasks, but I never felt like I was sabotaging anyone's ability to earn a living because I really did always focus on tasks that were truly repetitive and monotonous and not worthy of anyone's time. The whole "free people up to be more creative and strategic" isn’t practical or real, it's just hyperbole. So yes, I agree with you that we need concrete examples, properly contextualized, and some empathy for how automations might impact different people at different levels. And also, let's not use agents for things like writing emails and slacks to other humans -- they'll respond by using their own agents and to what end? It'll only prove all communication to be worthless.
Aside: on the Substack app I can't even find my way back to the original post. All I have is a link to this comment in my notes feed. Why is the Substack UI so bad? 😏
Good call-out, Jim. I’ve seen several modes of finding posts or Notes from comments on them. If I click a notification on my post, it only shows me that one comment thread (like this one), but I can scroll all the way to the top and there’s a nice ‘View post’ link. When it’s my comment on someone else’s post, there’s usually threading I can follow. But when I click on a notification of a Note comment, or a Note comment in my feed, the UI doesn’t seem to show a whole thread or Note. There’s no ‘View Note’ and, if the Note restacked a post, there’s no ‘View Post’. We have to tap, tap, tap to go up or down one level at a time, and it’s too easy to get lost. I don’t get why comments on Notes are treated differently from posts. It’s also a mixed blessing that comments on posts are sometimes (but not always) also posted as Notes, as duplicates. This part of the UI (and underlying data model) definitely could use some TLC!
"These agents excel at automating repetitive tasks, allowing human workers to focus on strategic activities that require creativity and critical thinking."
Curious, do you actually believe this? I see this exact phrasing used by so many AI proponents and always wonder what it means because not one person has been able to explain what knowledge works look like in the absence of the so-called repetative tasks. The thing is, in most white-collar jobs, strategic thinking and creativity are an occasional requirement, not the norm. I'm not saying this is okay, but it means AI agents, if they are as effective as purported, will definitely gouge a big hole in corporate America and the bleeding will be hard to stop.
Fair question, Jim. I’ll defer to Lakshmi as primary author to respond in depth.
My thought is that this statement is generally true of all automation — which has its limits. An agent that incorporates AI can potentially stretch the limits of what non-AI automation agents can do. It will still have its limits, not just in what it can do, but what it should be *allowed* to do unsupervised and without recourse.
And the humans who design automations that use AI agents need to be even more sensitive to what those limits are, and what new or worsened risks the use of AI introduces, and they must design and test and monitor accordingly. There are still ‘no silver bullets’.
I agree with you that it would be helpful to pick specific white-collar jobs, break down the everyday tasks, look at where automation might be useful, and assess whether AI is needed & what it contributes & where it falls short. We may find that there are nuances to a role that an AI agent’s training will not handle well. My own recent experiences with customer support bots bear this out.
One of my favorite quotes: “In theory, there is no difference between theory and practice. In practice, there is.”
Right now it seems too much like solutions in search of a problem. I've worked in tech for 20 years and have automated many tasks, but I never felt like I was sabotaging anyone's ability to earn a living because I really did always focus on tasks that were truly repetitive and monotonous and not worthy of anyone's time. The whole "free people up to be more creative and strategic" isn’t practical or real, it's just hyperbole. So yes, I agree with you that we need concrete examples, properly contextualized, and some empathy for how automations might impact different people at different levels. And also, let's not use agents for things like writing emails and slacks to other humans -- they'll respond by using their own agents and to what end? It'll only prove all communication to be worthless.
I couldn’t agree more, Jim. Thank you for sharing these thoughtful comments!
Jim, this post by Kenn So identifies jobs considered to be most amenable to AI-enhanced automation. The post has links to more detailed reports (I’m saving it to look into later). https://open.substack.com/pub/generational/p/ai-trends-2024?r=3ht54r
Aside: on the Substack app I can't even find my way back to the original post. All I have is a link to this comment in my notes feed. Why is the Substack UI so bad? 😏
Good call-out, Jim. I’ve seen several modes of finding posts or Notes from comments on them. If I click a notification on my post, it only shows me that one comment thread (like this one), but I can scroll all the way to the top and there’s a nice ‘View post’ link. When it’s my comment on someone else’s post, there’s usually threading I can follow. But when I click on a notification of a Note comment, or a Note comment in my feed, the UI doesn’t seem to show a whole thread or Note. There’s no ‘View Note’ and, if the Note restacked a post, there’s no ‘View Post’. We have to tap, tap, tap to go up or down one level at a time, and it’s too easy to get lost. I don’t get why comments on Notes are treated differently from posts. It’s also a mixed blessing that comments on posts are sometimes (but not always) also posted as Notes, as duplicates. This part of the UI (and underlying data model) definitely could use some TLC!