Incredibly, the list of activities for which I actively search out an AI tool is short: As a Product Owner, my main task is to facilitate communications, translate ideas, plan and keep schedules. While AI supports those tasks, I have been unable to offload most of my work, even though I really tried. And believe me, we tried.
3 ways how I use AI in my daily worklife
Yes, we’re actively experimenting with different AI solutions as part of our AI strategy. Beyond that, I also use AI in my daily work.
Research
First of all: never trust what AI tells you. AI excels at predicting the next word in any given sentence; it does not check whether that word conveys any truth. Still, knowing that risk, I use AI for research.
Instead of simply trusting my AI bot of choice (For Infodation that’s Copilot 365) to provide me with the correct answer, I investigate the sources that the LLM provides me when answering my prompt. I can then use my knowledge on media literacy to determine whether I think the source is trustworthy. AI provides me with a jumping off point for my research, from which I can find my way towards solid information.
Rewrite emails
Writing does not come easy for me. Business emails have the added complexity of having to contain dense information at the level that the intended target can digest and having to guide their reader towards the desired emotional state. Crafting an email is time-consuming work.
I let AI write my emotionally complex emails for me. I know what I want to say, and I know in which tone I want it to be said. Instead of spending 20 min on painstakingly writing each sentence, I add my work background, the emails content and style choices (length, sentiment, level of expertise) into Copilot 365 and let it create the email. Later, I spent another 5 minutes editing the email. To make the email sound like it was written by you, you can also upload previous emails into the LLM conversation and prompt the AI to emulate your writing style.
Summarize meetings
I cannot always make it to all the meetings that I am invited to. Life happens. I still want to be informed regarding the talking points of said missed meeting. Here is my solution: I ask another colleague to activate the auto-transcript function in teams. Once I am available again, I can feed the transcript into Copilot and ask for the summary and main talking points of said meeting. For meeting in foreign languages, you can also auto-translate the summary immediately.
Limitations here are that depending on the length of the meeting, the character count of Copilot can reach its limit. Also occasionally, a coworkers holiday plans that were discussed in the meeting, they will be mentioned the meeting summary.
AI as support
To paraphrase my colleague Hai Cao: “Copilot365 should only ever be your copilot”. As humans, we still need to oversee our work and our responsibilities. However, AI will evolve and so will how we interact with it. I personally am looking forward to seeing what the next iterations of AI will bring. For example, at Infodation we recently started testing our PO-agent, an internal tool that acts as a personal assistant to Scrum masters and product owners. Maybe I can finally unload some of my work on them.
