Nhu Y Le Thi

I really worry about how much people on this platform are outsourcing their thinking to AI. If you ask AI why the ocean is red, it's not going to say, "the ocean isn't red you ding dong", it's going to explain some of the things that could cause the ocean to be red. It may say the ocean isn't usually red, it may say something unusual has to happen for the ocean to appear red, but it isn't going to explain to you why you're actually straight up wrong (unless you've instructed it to do so). This is really dangerous for a few reasons: 1. It's reinforcing some bad thinking. If you ask, "how can I respond to this comment on LinkedIn", it's not going to go through a process of evaluating the conversation and presenting both sides; it's going to give you a comment to win the argument. 2. It feeds confirmation bias. AI is programmed to be helpful, so unless you've prompted it otherwise, it will try to give you exactly what you asked for. If you ask for evidence to support an idea, you may genuinely want multiple perspectives or to know if there is credible evidence to back up that idea. But AI will assume you're strictly looking for evidence to defend that idea and that's what it will give you. 3. It feeds you misinformation. AI does not have the same threshold for quality of research that human researchers do. It considers a Reddit thread, or an article from a low quality conspiracy theory website to be fair game. It's not going to prioritize high quality sources, and it may even cherry pick a few pieces of information out of context in an attempt to meet your request. That's going to leave you feeling even more confident in your misinformation, and it's going to make it less likely that you find contrary evidence - something which DOES happen when we conduct our own research. On top of that, AI helps you package your misinformation so that it sounds more credible and polished which makes it easier for you to believe, and for the misinformation to spread. A couple of things you can do to circumvent this: 1. Use AI as a thought partner. Ask it explicitly to challenge your assumptions, to evaluate the quality of evidence or arguments, and to offer counterpoints. 2. Conduct your own research. There's a reason people spend years in college just learning how to do good research, and by many people outsourcing all of this to AI, we're losing that skill of gathering a broad set of information, evaluating the credibility of resources, etc. So make sure that even if you're using AI as one of your research tools, conduct some additional research as well, and fact check everything you get from it. 3. Be intentional about your prompting. AI is extremely literal - prompts that are biased or based in a false assumption will give you biased or false outputs. I love AI, but I also know part of using it is recognizing its limitations as a tool, and it's potential negative impacts, and I think that's increasingly being lost.

375

99 Comments

Keep It Simple & Straightfoward — Thanks for the KISS Reminder, Susanna Matheson-Kwok! We had Susanna drop by the Soprano Design Singapore office recently, and it was a great reminder of something that often gets overlooked: the KISS principle — Keep It Simple and Straightforward. I came across this article on Business.com (worth a read!) that explains why keeping things simple is actually smart for business Some key takeaways: ✅ Don’t overcomplicate your product — focus on what people really need ✅ Keep your offerings and pricing easy to understand ✅ Make sure your website, emails, and customer touchpoints are clear ✅ Internally, simplify processes so everyone knows what to do and why It’s easy to get caught up in adding more features, more options, more steps. But often, less really is more — for your team and your customers. 👉 Don’t forget to "KISS" 👋 *************************** 🔔 Follow Miki Kong 🦂 for career, leadership support, and life insights. ♻️ Repost if you found this helpful. 💬 What others say about your trusted right-hand woman for leaders: https://lnkd.in/gdS2f5Tn

36

10 Comments