> ## Content Index > Fetch the complete content index at: https://fomo.observer/llms.txt > Use this file to discover other available public pages before exploring further. # TRIIN HERTMANN: How your AI agent makes me feel - URL: https://fomo.observer/triin-hertmann-how-does-your-ai-agent-make-me-feel/ - Published: 2026-08-06T06:00:56.000Z - Updated: 2026-08-06T06:00:55.000Z - Author: Triin Hertmann - Tags: Dispatches, AI, Triin Hertmann Ever pour your heart and soul into a presentation or a project just to be told it was 'ok'? It is so deflating. Recently, I had to present in a large important meeting. I prepared well, wrote the text, including personal jokes, fun metaphors, professional bullet points and practiced my speech. I felt it went well, I believe I came across warm and informative… but there were very minimal reactions from the listeners, like they were not actually present. Then I realised – half of them were AI agents and assistants. It stunned me. How did it make me feel? At this moment I was confused and disappointed, I guess. Like my properly heartily prepared dinner was put into a blender and mixed into colorless mush, that nobody really wants to eat later anyway. I am sure that in those AI summaries that were sent to the original invitees none of the warmth and jokes came across the way they were delivered. > We are social creatures and connection makes us feel good, belonging and understood. This need is not going anywhere and now, when we use AI every day, we also need to be aware of how it makes other people feel. ### AI everywhere Increasingly, like most people, I find myself reading emails, documents and reports that are clearly AI-generated and, more importantly, almost completely untouched by the person who sent them. Of course, on paper, they look flawless, but also incredibly hard to digest and feel anything about them. Somewhere along the way the author forgot to do the one thing that still matters most: decide what actually deserves another person's attention. It is surprisingly draining. Instead of helping me understand something more quickly I have to separate the important info from the fluff, identify what is actually being proposed, check whether the facts make sense and often rewrite the argument in my own head before I can even respond. When that happens, the reaction in my head is quite simple: *Did you really think my time was less valuable than yours?* > As an angel investor, I receive investment seeking emails from startups every week. 95% of them today are AI generated. One startup who actually bothered to send me a clearly personal and ‘handwritten’ email to ask for a call, I replied, “respect for personally written email, happy to hear more about what you are building!” I ended up investing and hopefully have good clear human correspondence with them for many happy years to come. ### Workslop In many cases, when used carelessly, AI does not eliminate work, it simply transfers it to someone else. Harvard Business Review recently described this phenomenon with a nicely fitting term: **workslop**. It refers to AI-generated work that looks professional but quietly shifts the cognitive burden onto someone else. Instead of reducing effort, it relocates it. Actually, workslop existed even before AI. I had a coworker whose texts were always full of grammar mistakes, weird repetitions and poorly worded paragraphs. I had to spend hours correcting her documents before I dared to put my name under it and send it out. She simply said, “thanks for reviewing!” My eye still starts twitching when I think about it. At first glance workslop seems like a quality problem, but I don't think quality is actually the issue. I've read brilliant AI-assisted writing and terrible human writing. The difference is whether the person using AI respected the next person's time. Can we just agree that we ask ourselves this question before we send anything out: *Would I want to read this if someone else sent it to me?* ### The social evaluation penalty To build on that - are we able to control and actually understand how the use of AI changes the way other people perceive **me**? It is fascinating that nobody thinks less of someone for using Excel instead of a calculator or GPS instead of a paper map. But with AI, it turns out, it’s different. I wanted proof of my intuition and found a [study](https://www.pnas.org/doi/10.1073/pnas.2426766122?ref=fomo.observer), **Evidence of a social evaluation penalty for using AI,** that explored this question. Across four experiments involving more than 4,400 participants, researchers found that people who use AI at work worry about being perceived more negatively, and those fears are justified. AI users were judged as less competent, less diligent and less independent than people who completed the same work without AI assistance. They were also more likely to assume that AI users were lazy or simply less capable of doing the work themselves. What is the logic behind it? When we see someone produce excellent work, we instinctively try to explain where that performance came from. If we believe a person has heavily relied on technology, we shift the credit for the results to AI and try to evaluate how much of it we believe belongs to the human. The work may be exactly the same quality, but our judgement on the person changes. The researchers describe this as a **social evaluation penalty**. In other words, AI can increase your productivity while simultaneously decreasing how competent other people believe you are. I have this amazing friend who multitasks in many fields so is super busy all the time. She is smart, talented and caring. But she also had a really annoying AI assistant that sent random requests, barged into meetings and just made people say, “she should really restrain her AI assistant.” The frustrating feeling started to, unfortunately, mirror into her personal image, too. Now she has put her AI assistant on a shorter leash and things are getting normal. > But I wanted to dig a bit deeper. As a woman, I see all kinds of gaps and judgements in our lives. So I wanted to know: Is the AI social evaluation penalty also dependent on gender? The answer is quite blunt: yes. Forbes recently published an [article](https://www.forbes.com/sites/michelletravis/2026/06/25/women-who-use-ai-seen-as-incompetent-men-who-use-ai-seen-as-pragmatic?ref=fomo.observer), **Women Who Use AI Seen As Incompetent; Men Who Use AI Seen As Pragmatic**, that describes how even with identical results, women are facing more **competence penalties** than men. This article points out a study where engineers were evaluated on their code quality. They compared AI users with none-AI users and later the gender of a programmer was revealed. > While male AI users received 6% lower competence ratings than non-AI users, female AI users received 13% lower competence ratings, even though they all produced identical computer code. The gap is real and women feel it. That has also been seen as one of the reasons women tend to use AI less, more carefully and also worry about the judgment. So, we subconsciously assume people are somewhat lazier, less caring and stupider when it’s clear they are using AI for completing tasks. And that women, yet again, pay more of the price. Feels like we are living in this interesting and hopefully temporary cultural transition. In time we need to learn new signals and re-evaluate the ways we judge work that combines human and machine effort. But also be really aware how our output feels for other people. How the group perceives me is crucial for my own well-being and success. After all, we are social creatures with instincts developed in thousands of years that can’t be turned off in a couple of months. ## Sign up for Fomo.Observer Startup stories from Estonia. Free weekly newsletter. Subscribe Email sent! Check your inbox to complete your signup. No spam. Unsubscribe anytime.