{"id":4092378,"date":"2025-11-18T14:47:26","date_gmt":"2025-11-18T19:47:26","guid":{"rendered":"https:\/\/www.computerworld.com\/video\/4092378\/why-ai-upskilling-is-failing-and-how-you-can-fix-it.html"},"modified":"2025-11-18T14:47:26","modified_gmt":"2025-11-18T19:47:26","slug":"why-ai-upskilling-is-failing-and-how-you-can-fix-it","status":"publish","type":"video_episodes","link":"https:\/\/www.computerworld.com\/video\/4092378\/why-ai-upskilling-is-failing-and-how-you-can-fix-it.html","title":{"rendered":"Why AI upskilling is failing, and how you can fix it"},"content":{"rendered":"<div id=\"remove_no_follow\">\n\n\n\n\n\n<div class=\"transcript\" id=\"transcript\" role=\"region\" aria-label=\"Transcript\" data-transcript-blocks=\"40\">\n            <h2 class=\"transcript__title\" id=\"transcript\">Transcript<\/h2>\n            <div class=\"transcript__text\">\n                <div class=\"transcript__list\" role=\"list\"><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>00:00<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"0\" aria-label=\"Jump to 00:00 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"1\"><p class=\"transcript__paragraph\">Keith Shaw: As companies embrace AI, the big topic around employee usage was upskilling. But recent research suggests many AI projects are failing because upskilling is failing. Why is this happening? We&rsquo;re going to dig into the issues on this episode of Today in Tech. Hi, everybody.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>00:25<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"25\" aria-label=\"Jump to 00:25 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"7\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">Welcome to Today in Tech. I&rsquo;m Keith Shaw. Joining me on the show today is Yvette Brown, co-founder of Xpromos, which trains people on generative AI and AI fluency. Welcome to the show, Yvette. Yvette Brown: Thanks, Keith. Happy to be here.<\/p><p class=\"transcript__paragraph\">Keith: There was recent research from MIT and others suggesting many AI projects are failing because employees lack the skills to run, deploy, or operate them. With upskilling such a big topic, why aren&rsquo;t employees getting the skills they need?<\/p><p class=\"transcript__paragraph\">Yvette: I&rsquo;m not sure I can answer why they aren&rsquo;t getting the skills, but it makes sense that a lack of skills is a big part of the problem. We often treat AI like a magic &ldquo;easy&rdquo; button, and it&rsquo;s not.<\/p><p class=\"transcript__paragraph\">Companies sometimes assume one person can be &ldquo;Debbie, the AI gal,&rdquo; and that&rsquo;s not how this works. This is like basic computer literacy &mdash; everyone needs to understand the tech stack if you want value from it. Many organizations still view this as something for engineers or tech teams.<\/p><p class=\"transcript__paragraph\">In reality, to get good results, everyday knowledge workers need to be skilled up on how to use these tools. Keith: You mentioned people treating AI as an easy button.<\/p><p class=\"transcript__paragraph\">We heard from another fully fluent guest who said the same thing &mdash; there&rsquo;s an assumption that AI will do everything you ask. From what you&rsquo;ve seen, what are the big mistakes people make when they take your courses or when you talk to them about AI fluency?<\/p><p class=\"transcript__paragraph\">Yvette: I&rsquo;ve been in marketing my entire career, and we have a saying about client briefs: &ldquo;Garbage in, garbage out.&rdquo;<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>02:33<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"153\" aria-label=\"Jump to 02:33 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"2\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">Yvette: As a marketer, you take what the client gives you and do your best &mdash; making a lot of assumptions because they don&rsquo;t fill it out completely. AI is similar.<\/p><p class=\"transcript__paragraph\">People say &ldquo;make me a plan&rdquo; or &ldquo;write me a blog&rdquo; without providing context, clarity, or constraints &mdash; no tone, no brand guidance, no details.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>02:43<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"163\" aria-label=\"Jump to 02:43 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"3\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">Yvette: We also fixate on the remaining time, not the time saved. If a task took six hours and now takes 30 minutes, you gained five and a half hours. Yes, you may still need to edit, but you&rsquo;re working faster &mdash; and often better &mdash; than before AI.<\/p><p class=\"transcript__paragraph\">We&rsquo;re not embracing that win. Keith: It feels like we&rsquo;re still in the copy-paste era: take the answer, paste it, and call it done. That&rsquo;s where people &ldquo;save&rdquo; those five hours &mdash; but it creates what people are calling AI &ldquo;work slop.&rdquo; Yvette: One hundred percent.<\/p><p class=\"transcript__paragraph\">Work slop is real, and I&rsquo;m glad it&rsquo;s getting surfaced. Your AI output doesn&rsquo;t have to be like that. In our training, an early lesson is: don&rsquo;t take the first answer AI gives you.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>05:00<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"300\" aria-label=\"Jump to 05:00 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"2\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">Yvette: There&rsquo;s no reason to accept the first draft. Take 30 seconds and ask it to look again. Don&rsquo;t just check for bias and factual accuracy &mdash; adjust the tone, push it to do better for your use case.<\/p><p class=\"transcript__paragraph\">With a few iterations, you&rsquo;ll get a tighter, better answer, more focused on what you&rsquo;re trying to communicate.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>05:31<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"331\" aria-label=\"Jump to 05:31 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"1\"><p class=\"transcript__paragraph\">Yvette: Part of the problem goes back to &ldquo;garbage in, garbage out.&rdquo; People don&rsquo;t always know what they want. As Brian Tracy says, &ldquo;Fuzzy goals get you fuzzy results.&rdquo;<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>05:53<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"353\" aria-label=\"Jump to 05:53 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"1\"><p class=\"transcript__paragraph\">Keith: Is there a difference between asking very specific questions versus general ones? When I do image generation, if I start with a huge, very specific prompt, the model sometimes loses details. I&rsquo;ll start generic and work toward specific. In your experience, is that better?<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>06:06<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"366\" aria-label=\"Jump to 06:06 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"2\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">Yvette: There&rsquo;s nuance and a kind of hierarchy. It&rsquo;s good to start with &ldquo;act as an expert in X,&rdquo; and the more specific, the better. Instead of &ldquo;act as a medical expert,&rdquo; say &ldquo;act as the leading thoracic surgeon at Cedars-Sinai,&rdquo; for example.<\/p><p class=\"transcript__paragraph\">That puts the LLM in the right headspace to answer. Hierarchical framing helps it retain details because, yes, it can lose its way.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>06:25<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"385\" aria-label=\"Jump to 06:25 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"1\"><p class=\"transcript__paragraph\">Yvette: One of my side hustles is prompt engineering for a couple of the big LLMs. A way to stump models is simply by changing the order of requests: they get it right once, then get confused the second time.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>07:14<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"434\" aria-label=\"Jump to 07:14 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"2\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">Keith: Are many people still in the question-and-answer mindset &mdash; treating AI like an advanced search engine? Yvette: Absolutely. Sometimes that&rsquo;s fine &mdash; if you want a simple fact with a few parameters.<\/p><p class=\"transcript__paragraph\">But for deeper work or multiple perspectives, you need more detail and a feedback loop: have a conversation.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>07:32<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"452\" aria-label=\"Jump to 07:32 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"3\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">Keith: Two tips you just gave: (1) after it answers, ask it to try again &mdash; especially for writing or production tasks; (2) ask the question a second way and explore beyond simple Q&amp;A. Yvette: Exactly. We joke it&rsquo;s like an international spice market: never take the first offer.<\/p><p class=\"transcript__paragraph\">Push back and you&rsquo;ll get better output. Keith: Anecdotally, I think a lot of people pick one tool &mdash; ChatGPT, Gemini, or Claude &mdash; and stick with it. Is there value in exploring multiple tools, or is one enough? Yvette: It depends.<\/p><p class=\"transcript__paragraph\">When you&rsquo;re learning and building AI fluency, one model is fine &mdash; it &ldquo;knows&rdquo; your tone and history, so it&rsquo;s easy to reference prior work.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>10:00<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"600\" aria-label=\"Jump to 10:00 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"2\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">Yvette: OpenAI has a huge market share, but there are many models. A friend of mine has hundreds of tabs open just for testing.<\/p><p class=\"transcript__paragraph\">Once you&rsquo;re fluent, you can compare models more effectively because you know what to look for &mdash; like testing fishing rods, you don&rsquo;t know what you don&rsquo;t know at first. Start with one, then expand.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>13:16<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"796\" aria-label=\"Jump to 13:16 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"2\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">Yvette: Also, we don&rsquo;t know where the economics of the tech giants land. If you build deep workflows in one model and it disappears or drastically changes pricing, you&rsquo;re exposed. In business, hedge your bets. Once you have a foundation, create Plan Bs and test alternatives to protect yourself.<\/p><p class=\"transcript__paragraph\">Keith: On upskilling: are your clients mostly individuals trying to improve themselves, or companies investing in teams? Yvette: Mostly companies &mdash; especially mid-size organizations, which are more nimble than some enterprises.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>14:44<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"884\" aria-label=\"Jump to 14:44 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"1\"><p class=\"transcript__paragraph\">Yvette: Individuals are interested too. We sometimes run cohorts through trade associations or networking groups. The advantage of a single-org cohort is collaboration: 10 people from the same company share learnings, which benefits the organization.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>15:00<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"900\" aria-label=\"Jump to 15:00 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"2\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">Keith: From a writing background, I rarely use AI to write emails &mdash; I can nail that task. Others struggle and rely on AI, and that&rsquo;s where a lot of &ldquo;work slop&rdquo; shows up. Any recommendations for using AI for email so it doesn&rsquo;t look like AI?<\/p><p class=\"transcript__paragraph\">Yvette: I also have a writing background. I usually write a first draft myself, then ask AI: &ldquo;Here&rsquo;s what I&rsquo;m trying to do &mdash; how can this be better?&rdquo; It suggests improvements, I select a few, and iterate.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>16:31<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"991\" aria-label=\"Jump to 16:31 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"2\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">Yvette: For better emails, be clear about your audience and objective. If you need to say, &ldquo;This is a bad idea&rdquo; in a corporate, non-hurtful way, tell the model that. Also, ask AI what signals make text sound AI-generated; it can list common giveaways and help filter them out.<\/p><p class=\"transcript__paragraph\">We even built a small GPT to help version thought-leadership pieces for different verticals &mdash; not to write them, but to customize language and then strip out the &ldquo;AI-ish&rdquo; tells.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>18:56<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"1136\" aria-label=\"Jump to 18:56 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"2\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">Keith: I once had ChatGPT &ldquo;improve&rdquo; a LinkedIn article I&rsquo;d written, and I hated the result &mdash; it didn&rsquo;t capture my tone. I&rsquo;ll use AI for outlines or key points, but I can usually write a better email myself. The people generating work slop often seem like buzzword lovers.<\/p><p class=\"transcript__paragraph\">Yvette: If you give nuanced direction &mdash; target audience, desired reaction, outcomes &mdash; it can help. Also, not everyone writes naturally well. For people who can articulate intent but struggle with tone, AI can be useful.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>19:03<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"1143\" aria-label=\"Jump to 19:03 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"2\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">Yvette: I had a client early on who said she wasn&rsquo;t very empathetic. A friend had a mishap, and she wanted to send a thoughtful note. AI helped her express what she felt.<\/p><p class=\"transcript__paragraph\">How you use AI varies by person &mdash; its purpose is to amplify your skills and fill gaps, not be everything to everyone.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>20:00<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"1200\" aria-label=\"Jump to 20:00 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"1\"><p class=\"transcript__paragraph\">Keith: It&rsquo;s interesting because AI often isn&rsquo;t great at empathy or humor, which is why people say managers won&rsquo;t be replaced by AI. Yvette: Right. Short messages can still be guided. And for humor &mdash;<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>20:59<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"1259\" aria-label=\"Jump to 20:59 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"1\"><p class=\"transcript__paragraph\">Yvette: &nbsp;&mdash; I once worked on a project asking AI to write jokes. A year and a half ago, they were&hellip; not good.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>21:53<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"1313\" aria-label=\"Jump to 21:53 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"2\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">Keith: Let&rsquo;s talk about under-appreciated capabilities beyond &ldquo;write this&rdquo; or &ldquo;analyze that&rdquo; or &ldquo;draw a picture.&rdquo; Yvette: One underutilized use is problem-solving through expert lenses. Traditionally, you&rsquo;d read a book &mdash; say Russell Brunson on funnels &mdash; then adapt it. LLMs contain a lot of this world knowledge.<\/p><p class=\"transcript__paragraph\">You can pose a business problem &mdash; even a small one like &ldquo;we&rsquo;re not getting enough top-of-funnel leads&rdquo; &mdash; and ask AI to convene a panel of experts and present frameworks.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>22:57<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"1377\" aria-label=\"Jump to 22:57 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"1\"><p class=\"transcript__paragraph\">Yvette: You can then say, &ldquo;Given my business context, combine these perspectives into a plan,&rdquo; drill it into a 90-day roadmap, then weekly tasks and checklists. You can simulate an entire strategy session in a couple of hours.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>23:24<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"1404\" aria-label=\"Jump to 23:24 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"2\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">Keith: My concern with long LLM sessions is the &ldquo;rabbit hole&rdquo; effect &mdash; like YouTube, where you drift from your original intent. I worry about hallucinations or just losing the plot. Yvette: Fair. Some models (ChatGPT especially) suggest next steps aggressively. I often ignore those and keep the thread focused.<\/p><p class=\"transcript__paragraph\">Yes, sometimes you go deeper than you planned; step away, then come back and extract what matters.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>25:00<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"1500\" aria-label=\"Jump to 25:00 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"2\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">Yvette: Another underused capability is solving one-off life tasks. I had a 40-year-old sprinkler system I couldn&rsquo;t remember how to test. I snapped a photo and asked AI. It gave me a 30-second step list &mdash; worked immediately.<\/p><p class=\"transcript__paragraph\">Same with rarely used software tasks: &ldquo;How do I add a page in Elementor?&rdquo; You can save 30 minutes &mdash; and a lot of frustration.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>30:00<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"1800\" aria-label=\"Jump to 30:00 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"2\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">Yvette: When GPT-5 launched, Sam Altman mentioned only about 7% of users had tried Deep Research. That blew my mind &mdash; it&rsquo;s so good. Try it for anything where you want sourced depth: competitors, a potential revenue stream, a neighborhood, a vacation plan.<\/p><p class=\"transcript__paragraph\">It asks a few clarifying questions, works for a bit, and returns a thorough report with links &mdash; usually less hallucinatory than ad-hoc chat citations.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>30:26<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"1826\" aria-label=\"Jump to 30:26 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"1\"><p class=\"transcript__paragraph\">Keith: Maybe the name &ldquo;Deep Research&rdquo; scares people &mdash; sounds academic or &ldquo;for professors.&rdquo;<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>30:30<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"1830\" aria-label=\"Jump to 30:30 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"1\"><p class=\"transcript__paragraph\">Yvette: You&rsquo;re not wrong. Sometimes engineers name things, and, well&hellip; marketing doesn&rsquo;t always get a pass at the label.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>31:05<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"1865\" aria-label=\"Jump to 31:05 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"2\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">Keith: We&rsquo;re moving into agentic AI. Are we going to repeat the same mistakes from early generative AI? Can you safely use agents without basic fluency? Yvette: In my opinion, no. If you aren&rsquo;t AI-fluent, agents are much riskier than chat models.<\/p><p class=\"transcript__paragraph\">The opportunity for things to go south is larger.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>31:51<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"1911\" aria-label=\"Jump to 31:51 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"1\"><p class=\"transcript__paragraph\">Yvette: It&rsquo;s early days for agents. Playing with models in a contained way is one thing. But when you connect agents to databases and sensitive information &mdash; what could go wrong? A lot. Beyond model failure, you have cybersecurity risks.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>32:01<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"1921\" aria-label=\"Jump to 32:01 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"1\"><p class=\"transcript__paragraph\">Keith: Spoken like an IT security pro. People fear connecting agents to secrets. You don&rsquo;t want someone casually unleashing the KFC recipe. Yvette: Exactly.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>34:37<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"2077\" aria-label=\"Jump to 34:37 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"2\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">Keith: For someone looking to upskill on agents, what should they do &mdash; keep building general fluency first? Yvette: Yes. One path is learning to create an AI pilot.<\/p><p class=\"transcript__paragraph\">Treat it like product development: iterate, track versions, shadow the agent&rsquo;s output before going live, define pass\/fail criteria, and run test batches. Then, schedule periodic reviews to ensure it hasn&rsquo;t drifted or started hallucinating.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>35:00<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"2100\" aria-label=\"Jump to 35:00 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"2\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">Yvette: We imagine knowledge workers managing multiple agentic workflows &mdash; updating them as models change &mdash; rather than doing every deliverable manually. Keith: Have you seen learners go from clueless to confident by the end of your program? Yvette: Definitely.<\/p><p class=\"transcript__paragraph\">People arrive at different levels; early on, some think, &ldquo;Yeah, I already do that.&rdquo; As we progress, they lean in. By the end, most say, &ldquo;I didn&rsquo;t know what I didn&rsquo;t know.&rdquo;<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>36:26<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"2186\" aria-label=\"Jump to 36:26 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"1\"><p class=\"transcript__paragraph\">Yvette: Marketers, for example, learn to build an AI pilot and produce a Pilot Handbook &mdash; stepping toward a Center of Excellence. Many say, &ldquo;This is great, but I don&rsquo;t want to build it &mdash; at least now I know what to ask the tech team.&rdquo;<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>37:10<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"2230\" aria-label=\"Jump to 37:10 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"1\"><p class=\"transcript__paragraph\">Keith: Low-code\/no-code and &ldquo;citizen developers&rdquo; worry me. I once asked for a pizza-night scheduling system; it returned code and server steps &mdash; I was out. Yvette: And that&rsquo;s fine. AI is great for a concept or MVP you can show clients &mdash; better concept boards than we used to hand-sketch.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>38:24<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"2304\" aria-label=\"Jump to 38:24 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"1\"><p class=\"transcript__paragraph\">Yvette: As a marketer, I know interactive lead magnets outperform static PDFs. I built a 10-question interactive quiz with progress bar and outcomes in a couple of hours &mdash; something that used to take ~20 hours with developers. I&rsquo;m not a coder, but it was just enough HTML to ship.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>40:00<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"2400\" aria-label=\"Jump to 40:00 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"3\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">Keith: Give viewers an &ldquo;AI homework assignment&rdquo; to improve fluency &mdash; something practical beyond &ldquo;take a course.&rdquo; Yvette: First, show that AI already knows your job.<\/p><p class=\"transcript__paragraph\">Paste your LinkedIn profile or describe your role (&ldquo;I&rsquo;m X at a mid-size Y-type company&rdquo;) and ask it to list your likely deliverables and how it can help you do each better. Then ask it to prioritize those deliverables and estimate time savings.<\/p><p class=\"transcript__paragraph\">People are often shocked by how accurately it understands roles. Second, run one Deep Research on anything relevant to your life &mdash; college options, a neighborhood, a market landscape &mdash; just to experience the depth.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>42:03<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"2523\" aria-label=\"Jump to 42:03 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"2\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">Keith: We&rsquo;re entering the holiday shopping season. Many people are using AI to help with shopping. Could this be the consumer &ldquo;killer app&rdquo; or an agentic push? Yvette: Maybe. Starting with gift lists based on what you know about people is a great use case.<\/p><p class=\"transcript__paragraph\">Will full in-chat purchasing (e.g., Walmart integrations) become the thing? Not sure.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>42:54<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"2574\" aria-label=\"Jump to 42:54 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"1\"><p class=\"transcript__paragraph\">Yvette: Even if people learn one use (like shopping), they may not realize everything AI can do &mdash; like new phone features most of us never use. Keith: Exactly.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>43:06<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"2586\" aria-label=\"Jump to 43:06 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"2\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">I&rsquo;m often a few iPhone generations behind, so I miss features for years. I worry the same happens with AI &mdash; people using only 10% of what&rsquo;s possible. Yvette: I agree. That&rsquo;s why we&rsquo;re out here &mdash; like Johnny Appleseed &mdash; helping people understand the breadth of capability.<\/p><p class=\"transcript__paragraph\">AI can give you time back and improve quality so you can focus on higher-value work. It enhances the human experience &mdash; whether that&rsquo;s more time with prospects in sales or better creative outputs.<\/p><\/div><\/div><div class=\"transcript__item\" role=\"listitem\"><div class=\"transcript__item-meta\">\n                    <time>45:00<\/time> \n                    <button class=\"transcript__jump-button\" data-time=\"2700\" aria-label=\"Jump to 45:00 in video\">\n                        <i class=\"icon-play\"><svg><use xlink:href=\"#icon-play\"><\/use><\/svg><\/i>\n                    <\/button>\n                <\/div><div class=\"transcript__content\" data-paragraphs=\"3\" data-readmore=\"Show more\" data-readless=\"Show less\"><p class=\"transcript__paragraph\">Yvette: If you have a question about anything in life, ask AI and see where the conversation goes &mdash; you&rsquo;ll be surprised. Keith: And we got a Johnny Appleseed reference!<\/p><p class=\"transcript__paragraph\">I live near where he grew up &mdash; Leominster, Massachusetts &mdash; home of a big Johnny Appleseed festival, so I&rsquo;m thrilled. Yvette, thanks for the tips and insights. Yvette: Thanks for having me. Keith: That&rsquo;s all the time we have for this week&rsquo;s episode.<\/p><p class=\"transcript__paragraph\">Be sure to like the video, subscribe to the channel, and add your thoughts below if you&rsquo;re watching on YouTube. Join us every week for new episodes of Today in Tech. I&rsquo;m Keith Shaw &mdash; thanks for watching.<\/p><\/div><\/div><\/div>\n            <\/div>\n        <\/div>\n\n\n<p><\/p>\n\n\n<\/div>\n","protected":false},"author":1701,"featured_media":100069189,"template":"","meta":{"__idg_published_ids":[],"__idg_published_status":"draft","embargo_date":"","multi_title":"[]","old_id_in_onecms":"","_idg_updated_flag":false,"_idg_updated_date":"","hreflang_xdefault":0,"content_type":"","suppress_html_meta":"{}","byline":"","featured_video_id":0,"supress_floating_video":false,"prevent_index":0,"has_duration":0,"teaser_paragraphs":"","is_translated_post":0,"idg_original_post_id":0,"idg_translated_post_ids":[],"idg_original_post_publication":"","idg_original_post_language":"","idg_original_post_brand":"","right_panel_heading":"","video_type":"youtube","podcast_episode_id":0,"yt_video_id":"atr2QeuqubE","yt_video_duration":"46.18","jw_video_id":"","jw_video_duration":"","jw_video_title":"","episode_ordering":351,"is_hosts_enabled":false,"is_guests_enabled":false,"suppress_monetization":"{\"content_ads\":false}"},"categories":[1885,2888,1932,1933],"tags":[],"languages":[21],"editions":[12],"publication":[9,10],"sponsorships":[],"video_series":[6607],"coauthors":[],"class_list":{"0":"post-4092378","1":"video_episodes","2":"type-video_episodes","3":"status-publish","4":"has-post-thumbnail","6":"category-artificial-intelligence","7":"category-generative-ai","8":"category-it-skills","9":"category-it-training","10":"languages-en","11":"editions-global","12":"publication-computerworld","13":"publication-us-default","14":"video_series-today-in-tech"},"eyebrow":{"eyebrow":"Video","eyebrow_style":"default","eyebrow_feed_title":"Video","eyebrow_feed_style":"default"},"_links":{"self":[{"href":"https:\/\/www.computerworld.com\/wp-json\/wp\/v2\/video_episodes\/4092378","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.computerworld.com\/wp-json\/wp\/v2\/video_episodes"}],"about":[{"href":"https:\/\/www.computerworld.com\/wp-json\/wp\/v2\/types\/video_episodes"}],"author":[{"embeddable":true,"href":"https:\/\/www.computerworld.com\/wp-json\/wp\/v2\/users\/1701"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.computerworld.com\/wp-json\/wp\/v2\/media\/100069189"}],"wp:attachment":[{"href":"https:\/\/www.computerworld.com\/wp-json\/wp\/v2\/media?parent=4092378"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.computerworld.com\/wp-json\/wp\/v2\/categories?post=4092378"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.computerworld.com\/wp-json\/wp\/v2\/tags?post=4092378"},{"taxonomy":"languages","embeddable":true,"href":"https:\/\/www.computerworld.com\/wp-json\/wp\/v2\/languages?post=4092378"},{"taxonomy":"editions","embeddable":true,"href":"https:\/\/www.computerworld.com\/wp-json\/wp\/v2\/editions?post=4092378"},{"taxonomy":"publication","embeddable":true,"href":"https:\/\/www.computerworld.com\/wp-json\/wp\/v2\/publication?post=4092378"},{"taxonomy":"sponsorships","embeddable":true,"href":"https:\/\/www.computerworld.com\/wp-json\/wp\/v2\/sponsorships?post=4092378"},{"taxonomy":"video_series","embeddable":true,"href":"https:\/\/www.computerworld.com\/wp-json\/wp\/v2\/video_series?post=4092378"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/www.computerworld.com\/wp-json\/wp\/v2\/coauthors?post=4092378"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}