33 min
MC73 GenAI: Friend or Foe? Management Café
-
- Management
In 2024 Generative AI is still an a rapidly emerging technology. We know it's going to impact lots of jobs and work behaviour. But it's not yet apparent how that impact will be felt.
There is a big difference between ChatGPT helping you with a task (e.g. summarising your notes or drafting a communication) vs using it to entirely perform the task (e.g. using AI generated output without editing). Managers need to be mindful of how they are using AI, especially as it is increasingly embedded in digital workplace communication tools.
Transparency is also key - people should know why something might sound a little different than your normal style. Ultimately with a technology as prevalent and seductive as Gen AI, it's critical that we use it as an aid. We must not lose ourselves or our human connection to those we manage and work with.
Show notes:
02:00 mins Our ability to experiment with new technology is much influenced by our context. Freelancers or those with lots of autonomy have an easier time experimenting with new tech. But those working in large, heavily controlled environments might have to wait a lot longer before they get to play.
3:45 AI is also showing up in software tools for example Helpscout bought an AI company to automatically draft replies to inbound support requests.
4:25 Kona have created an AI powered management coach in Slack. Tim is very uneasy about this - he feels that an authentic response, even if it isn't perfect, is still better than a "best practice" response. But Pilar rightly points out that some managers will just google "five tips for how to..." and then follow that advice. Is AI any worse?
6:15 AI can be even easier to ask than a human, e.g. if you had an external support network like we discussed in #65. AI is always available and quite forgiving. You don't even have to say please or thank you. Pilar has had to guard against becoming lazy because of the help she can get from ChatGPT.
8:10 Tim takes a "get off my lawn" approach that hard things, like writing a performance review, should be hard. Struggling through them makes you better at them and that experience is valuable. He also worries about the message it sends to your employees if they received a performance review written by ChatGPT.
10:00 Pilar outlines a different scenario where someone who has all the skills of a great manager but struggles to produce a well structured written summary. ChatGPT can help that manager draft the document. If the manager should be transparent with the employee that ChatGPT was used in creating the summary, the result can be a good one. After all back in the day a senior manager might have gotten their secretary to write up a performance review for them based on notes provided.
13:45 To get the best outcome the input needs to be detailed and high quality. And the final product must be edited and reviewed by the creator so that it reflects the reality of what they were trying to communicate.
16:15 Mindful use of generative AI takes a bit of work, but it means that you really are using it to represent your own thoughts and ideas.
17:55 The key for Pilar is transparency. Creators should disclose where they have used AI.
19:30 A danger with ChatGPT is it can disincentivise us from getting better at some valuable skills.
20:30 Rapidly emerging technology can create issues between those who are onboard with it and those who are not. People's awareness and tolerance and adoption happen at different paces.
21:30 Tim is concerned that use of AI by managers can damage authenticity and connection. It takes conscious effort to maintain human connection in a digital workplace, technology doesn't provide this by default. And yet AI is increasingly embedded in our communication tools.
23:00 Pilar agrees, but advocates for the benefits of mindful use. For example, someone who isn't very proficient in English could get great benefits from ChatGPT helping th
In 2024 Generative AI is still an a rapidly emerging technology. We know it's going to impact lots of jobs and work behaviour. But it's not yet apparent how that impact will be felt.
There is a big difference between ChatGPT helping you with a task (e.g. summarising your notes or drafting a communication) vs using it to entirely perform the task (e.g. using AI generated output without editing). Managers need to be mindful of how they are using AI, especially as it is increasingly embedded in digital workplace communication tools.
Transparency is also key - people should know why something might sound a little different than your normal style. Ultimately with a technology as prevalent and seductive as Gen AI, it's critical that we use it as an aid. We must not lose ourselves or our human connection to those we manage and work with.
Show notes:
02:00 mins Our ability to experiment with new technology is much influenced by our context. Freelancers or those with lots of autonomy have an easier time experimenting with new tech. But those working in large, heavily controlled environments might have to wait a lot longer before they get to play.
3:45 AI is also showing up in software tools for example Helpscout bought an AI company to automatically draft replies to inbound support requests.
4:25 Kona have created an AI powered management coach in Slack. Tim is very uneasy about this - he feels that an authentic response, even if it isn't perfect, is still better than a "best practice" response. But Pilar rightly points out that some managers will just google "five tips for how to..." and then follow that advice. Is AI any worse?
6:15 AI can be even easier to ask than a human, e.g. if you had an external support network like we discussed in #65. AI is always available and quite forgiving. You don't even have to say please or thank you. Pilar has had to guard against becoming lazy because of the help she can get from ChatGPT.
8:10 Tim takes a "get off my lawn" approach that hard things, like writing a performance review, should be hard. Struggling through them makes you better at them and that experience is valuable. He also worries about the message it sends to your employees if they received a performance review written by ChatGPT.
10:00 Pilar outlines a different scenario where someone who has all the skills of a great manager but struggles to produce a well structured written summary. ChatGPT can help that manager draft the document. If the manager should be transparent with the employee that ChatGPT was used in creating the summary, the result can be a good one. After all back in the day a senior manager might have gotten their secretary to write up a performance review for them based on notes provided.
13:45 To get the best outcome the input needs to be detailed and high quality. And the final product must be edited and reviewed by the creator so that it reflects the reality of what they were trying to communicate.
16:15 Mindful use of generative AI takes a bit of work, but it means that you really are using it to represent your own thoughts and ideas.
17:55 The key for Pilar is transparency. Creators should disclose where they have used AI.
19:30 A danger with ChatGPT is it can disincentivise us from getting better at some valuable skills.
20:30 Rapidly emerging technology can create issues between those who are onboard with it and those who are not. People's awareness and tolerance and adoption happen at different paces.
21:30 Tim is concerned that use of AI by managers can damage authenticity and connection. It takes conscious effort to maintain human connection in a digital workplace, technology doesn't provide this by default. And yet AI is increasingly embedded in our communication tools.
23:00 Pilar agrees, but advocates for the benefits of mindful use. For example, someone who isn't very proficient in English could get great benefits from ChatGPT helping th
33 min