This ONE AI Agent Will Make You So Productive It's Almost Unfair — Transcript

Full transcript

1. 0:00I have spent many years as a CEO and
2. 0:03board member in tech [music] companies
3. 0:05and AI companies and here's what I am
4. 0:08seeing. One AI tool that knows you
5. 0:11[music] really well is better than 10
6. 0:13that don't. And the longer you use it,
7. 0:15the more productive you become. So, in
8. 0:18this video podcast, I will show you how
9. 0:20to build that one personal AI agent,
10. 0:24whichever [music] platform you use, and
11. 0:26we will use that personal AI agent to
12. 0:28defeat the four forces that drain your
13. 0:31productivity. The first three forces
14. 0:32[music]
15. 0:33steal your hours, but the last one
16. 0:37steals your years. The best AI tool is
17. 0:40the one that saves you from opening
18. 0:42another one. Whatever you're trying to
19. 0:44accomplish, [music] your time and your
20. 0:46attention are limited by default. The
21. 0:48more you spend managing the little
22. 0:51things, the less you'll have left for
23. 0:53the [music] things that matter. So,
24. 0:55here's the challenge. Instead of one
25. 0:58great AI chief of staff, we end up
26. 1:00hiring a crowd of AI interns, and you
27. 1:04end up spending more time managing them,
28. 1:06and you become their chief of staff, and
29. 1:09that needs to change. So, what makes
30. 1:12these personal AI agents different from
31. 1:15the chatbots that you've already used?
32. 1:17Well, there are three things. How do you
33. 1:19remember them? The easiest way to think
34. 1:21about it is what I call act.
35. 1:24First, they're autonomous. So, you give
36. 1:27it a goal and it figures out how to get
37. 1:30the job done. It runs on its own
38. 1:33computer in the cloud. It can work with
39. 1:35other tools, other agents, other apps.
40. 1:38It can browse the web. It can take
41. 1:40actions. Second, it's continuous. So, it
42. 1:44doesn't stop working when you close the
43. 1:47tab or you close the chat or you close
44. 1:50your laptop. It can orchestrate a task
45. 1:52that can run for hours. It'll come back
46. 1:55to you when it's done. And third,
47. 1:58tailored. It is crafted and trained with
48. 2:01personalization in mind. So, it
49. 2:03remembers your goals, your preferences,
50. 2:06and how you like things done. And by the
51. 2:10way, I'm going to use Cache PT's DOTS,
52. 2:13which is their personal AI agent for the
53. 2:15demos in this video, but the principles
54. 2:18apply to all personal AI agents, whether
55. 2:21you're using Claude Co-work or Google's
56. 2:25Gemini agent or Meta has Muse now or
57. 2:28Grock has Grockbot. So you can pick your
58. 2:31favorite tool. The approach is the same.
59. 2:35Now one rule to remember, let them
60. 2:37manage, let them execute, let them
61. 2:39partner with you, but it's you who makes
62. 2:42the final call. Otherwise, your agent
63. 2:46might end up posting your bank statement
64. 2:48on your Instagram story. So be careful.
65. 2:52Let's start with the first force,
66. 2:54fragments. There is a job you do 1,200
67. 2:58times a day and it's not in your job
68. 3:01description. Sometimes all these AI
69. 3:04tools make me feel like I am a tool. In
70. 3:082022,
71. 3:09researchers published an article in
72. 3:12Harvard Business Review. They tracked
73. 3:14employees of Fortune 500 companies over
74. 3:17a month and they tracked and logged
75. 3:20every time someone switched from one app
76. 3:23or website to another. The average
77. 3:26person switched nearly 1,200 times a
78. 3:30day. Now each switch costs only about 2
79. 3:34seconds. So that's about 40 minutes a
80. 3:36day. But that's as long as a lunch
81. 3:39break. But there's something more
82. 3:41alarming in terms of hidden cost. So
83. 3:44fragmentation damages the the quality of
84. 3:47your attention. Why? Because the
85. 3:50information we need is scattered
86. 3:52everywhere. So you spend time hunting
87. 3:56for the information you need to do the
88. 3:59work before you can actually do the
89. 4:01work. That's the first force. Your
90. 4:03personal AI agent has to fight. that
91. 4:05force and that AI agent needs to become
92. 4:08the glue that connects all your apps and
93. 4:11information islands so you can focus on
94. 4:14harvesting the info not hunting for it.
95. 4:17Let's go through an example. Imagine you
96. 4:20have landed an interview for a job you
97. 4:23have wanted for years. This could change
98. 4:26your career. But now what? Because your
99. 4:29resume is in Google Drive somewhere.
100. 4:32Your notes from previous interviews sit
101. 4:34somewhere else. The job description is
102. 4:36somewhere else. Your conversations with
103. 4:38the recruiter are in some emails.
104. 4:41[music] The company background is out
105. 4:43there somewhere. The evidence of your
106. 4:45best work is actually scattered across
107. 4:48the entire drive in presentations and
108. 4:51documents and emails. [music] So, this
109. 4:53would take days to consolidate. How do
110. 4:56you position yourself for this
111. 4:58interview? Why not try something
112. 5:00different?
113. 5:01You can go for a walk with your personal
114. 5:05AI agent. You can prompt it. You can
115. 5:08say, "I've got an interview for a job I
116. 5:12have wanted for years. Help me walk in
117. 5:16as the most prepared candidate they've
118. 5:19ever seen. Use my resume, my past work,
119. 5:22my emails, company background, and
120. 5:25research the company's interview
121. 5:27process." Now, this could be any
122. 5:29company, but here's what companies are
123. 5:32doing right now. McKenzie, for example,
124. 5:35is testing AI agents that play a client
125. 5:39in interviews for entry-level
126. 5:41consultants. So, the candidate has to
127. 5:44work through the client's problem live
128. 5:47while the human interviewer is watching
129. 5:49and scoring them. That's the world we're
130. 5:52walking into. So, your interview prep
131. 5:55has to move up a level, too, right? You
132. 5:57tell your AI agent to simulate that
133. 6:00interview. The prompt would be something
134. 6:02like, "Now simulate the interview based
135. 6:04on what you've learned about me and the
136. 6:07role and the company. Ask me realistic
137. 6:10questions, challenge my answers, and
138. 6:13push back when my reasoning is weak.
139. 6:16Don't make it easy." And then when
140. 6:18you're finished, you can ask it to
141. 6:20critique your performance, your answers.
142. 6:22You can write up notes, and you can do
143. 6:25it again. Now, here's the best part. If
144. 6:28the interview goes badly, hey, who
145. 6:31cares? That was just a rehearsal. Do it
146. 6:34again and again. At least you got your
147. 6:3610,000 steps in. So, your agent gathers
148. 6:40everything. It visits seven different
149. 6:43tools and apps and acts like your
150. 6:45interviewer. And you focus on just one
151. 6:48thing, getting better and getting that
152. 6:50job. Now remember that these agents can
153. 6:53get information from other places only
154. 6:56after you connect them to those specific
155. 6:59apps and tools. Now let's move to
156. 7:02fighting the second force that slows
157. 7:04down productivity flood. 117 emails, 153
158. 7:10chat messages every single day. The best
159. 7:14thing your AI agent can do might be what
160. 7:17it never shows you. In 2025, Microsoft
161. 7:21surveyed about 31,000 people. The
162. 7:24average worker [music] received 117
163. 7:28emails and 153 Teams messages a day.
164. 7:34Most of those emails got less than a
165. 7:36minute of attention, sure, but that's
166. 7:39the flood of information. And that's not
167. 7:42even including the social media stuff
168. 7:44that comes at you from every direction.
169. 7:46And somewhere in there is that one
170. 7:49message that actually mattered. [music]
171. 7:52And also the things that matter are
172. 7:55usually pretty complex. So researchers
173. 7:58at Carnegie Melon tested Americans with
174. 8:01private insurance on four basic terms
175. 8:05[music] like deductible and co-ay. And
176. 8:09only 14% [music] got all four
177. 8:12definitions right. So you're fighting
178. 8:14two floods really. Too much reaches you
179. 8:16all the time and at times it is too
180. 8:19dense [music] to act on. And this is
181. 8:22where your personal AI agent comes in
182. 8:24handy. [music] It can filter the noise
183. 8:27and decide what deserves your attention
184. 8:29and simplify what deserves your
185. 8:31judgment. And that's exactly what a
186. 8:34chief of staff would do. Right? Of
187. 8:37course that's also the risk. Whoever or
188. 8:41whatever is the gatekeeper decides what
189. 8:44you get to see. So don't stop at setting
190. 8:48the goals. Check the work and keep
191. 8:51refining. So let's create that filter.
192. 8:53Prompt your personal AI agent something
193. 8:55like I want to protect my focus and I
194. 8:59want to only get pulled away for what
195. 9:01truly needs me. So, at 10:00 a.m., 100
196. 9:05p.m., and 400 p.m., check my emails, my
197. 9:08messages, my calendar, and flag only
198. 9:12decisions waiting for me. Deadlines that
199. 9:15moved or changes that I need to know
200. 9:18about, park everything else for a 5:00
201. 9:21p.m. summary. And this is a great
202. 9:24prompt. I use it all the time because
203. 9:26you'll be surprised how much of your
204. 9:29urgent work can wait until 5:00 p.m. and
205. 9:33some of it can even wait until you
206. 9:35retire. And okay, what about the
207. 9:38complexity? Well, this is how you
208. 9:40untangle it. Let's say you're reviewing
209. 9:44the lease for your new apartment. Okay,
210. 9:47get it and give it to your personal
211. 9:49agent and it will help. You can say
212. 9:50something like, "I'm about to sign this
213. 9:53lease and I want to know exactly what
214. 9:56I'm agreeing to." You see, in all of
215. 9:59these prompts, you're giving objectives,
216. 10:02not just instructions, goals, not just
217. 10:05guidance. Flag anything that could cost
218. 10:07me money or limit what I can do, and
219. 10:11show me the exact clauses. Great prompt.
220. 10:15You now have a translation layer between
221. 10:17complexity and simplicity. You can
222. 10:21quickly move from consuming information
223. 10:23to making decisions. And that brings us
224. 10:26to the third force, frequency. The work
225. 10:30that keeps coming back whether you're
226. 10:33ready or not. You remember to eat dinner
227. 10:36every night. It's a habit. You'll
228. 10:38remember you're flying to Paris with the
229. 10:42love of your life 3 weeks from now and
230. 10:45you can't wait to start packing. You're
231. 10:46counting days. It's a headline event.
232. 10:50But why do you have trouble remembering
233. 10:52to take that vitamin once a week? I do
234. 10:57for two reasons. First, a habit is built
235. 11:00by repetition. And a weekly task doesn't
236. 11:04get enough repetition to become a habit.
237. 11:07Second, we all need a trigger to remind
238. 11:10us about those infrequent but recurring
239. 11:13tasks. No habit, no headline, no hope of
240. 11:17tracking it. Unless, of course, you have
241. 11:20your friendly neighborhood AI agent. Let
242. 11:23me give you a a real example.
243. 11:26A few weeks ago, I bought my iPhone 18.
244. 11:31You know, I usually buy my iPhone every
245. 11:33four years or so.
246. 11:36There's no reason to upgrade it
247. 11:37otherwise. But 2 days ago, my agent
248. 11:40reminded me, "Hey, your trade-in kit
249. 11:43arrived about 14 days ago. So today is
250. 11:46the last day for you to return your old
251. 11:48iPhone for your trade in." I was like,
252. 11:52"Oh, wow. Thank you, Capes." That's her
253. 11:55name, by the way. In my mind, she's a
254. 11:58superhero who can juggle a hundred
255. 12:00things at once. and I can't remember
256. 12:04where I left my phone. So, here's
257. 12:06another example. My jaw used to block my
258. 12:09air waves when I used to sleep. So, I
259. 12:12would wake up 34 times every hour
260. 12:16because my brain would jolt me awake
261. 12:19every time I stopped breathing. So, it's
262. 12:22it's fixed now. Uh, but the other day I
263. 12:26sent Capes a quick line saying, "When
264. 12:28was the last time I saw my
265. 12:30pulmonologist, the sleep disorder
266. 12:33specialist?" And of course, she came
267. 12:34back and said, "It's been a year." So, I
268. 12:38asked her to set up an appointment. And
269. 12:40then, this is the most important thing.
270. 12:43I said, "Make sure I see my doctor every
271. 12:473 to four months." And that's how you
272. 12:50make your agent your trigger. And here's
273. 12:53another thing to keep in mind. My AI
274. 12:56agent only set the appointment. It did
275. 12:59not have my health records or any
276. 13:02financial information. Now, I'm not
277. 13:05comfortable sharing any of that with AI.
278. 13:08That's my choice for now. You can make
279. 13:10your own judgment. So, here's your
280. 13:12action item. Pick one thing in your life
281. 13:15with no habit and no headline and hand
282. 13:19it to your agent this week. And by the
283. 13:21way, if you like these ideas, you'll
284. 13:23enjoy my newsletter as well. It goes out
285. 13:26every Tuesday. One inside, one tool, one
286. 13:30practice, and it's free. I love writing
287. 13:33it. The link is right below. So far,
288. 13:35these agents are helping you earn back
289. 13:38your attention, your evenings, your
290. 13:40peace of mind. But that leads us to a
291. 13:44bigger question. What are you going to
292. 13:46do with it? That's the fourth force,
293. 13:49future. The first three forces steal
294. 13:52your hours. This one [music]
295. 13:55steals your years. I'll do it later. Is
296. 13:58how life slips [music] away. There is a
297. 14:02famous video on YouTube called the last
298. 14:06[music] lecture. You may have seen it
299. 14:09already. It's by computer science
300. 14:12professor Randy Pouch and he gave it
301. 14:16many years ago when he knew that he only
302. 14:20had a few months to live. [music] He had
303. 14:23pancreatic cancer and um he also gave
304. 14:27another lecture before passing away and
305. 14:31that was on time [music] management. In
306. 14:34that one, he talks about how Americans
307. 14:39as a culture [music] are very interested
308. 14:42in money, how much someone earns, how
309. 14:46they spend it, their net worth, and so
310. 14:49on and on. But we're not that interested
311. 14:52in time.
312. 14:54We make a budget on how we spend money,
313. 14:58but we don't budget how we spend time.
314. 15:01And Ry's insight was
315. 15:04we have it backwards. He passed away
316. 15:07less than a year after that lecture. He
317. 15:10was 47 and [clears throat] he knew the
318. 15:14value of time. So he spent it on what
319. 15:17mattered the most. So here's the real
320. 15:20question. AI agents will give you back
321. 15:23time. What are you going to do with it?
322. 15:25Productivity is great only when it helps
323. 15:28you achieve your purpose. So, here's the
324. 15:32action item. Write down three things you
325. 15:36keep saving for the future. And then
326. 15:38tell your agent, "These are the
327. 15:40priorities. Every Sunday evening, review
328. 15:43my calendar and work from this week.
329. 15:47Tell me what I did to achieve these
330. 15:50priorities and what I should change next
331. 15:52week." Now your agent moves you to what
332. 15:57matters to you and toward your future.
333. 16:00The strange thing about time is that it
334. 16:05feels precious only when there is less
335. 16:09of it left.
336. 16:11So take a look at your watch. Really
337. 16:15look.
338. 16:16You know what time it is? It's your
339. 16:19time.
340. 16:22Thank you. And I love you.