Latest / Roaming Returns / 158 - I Asked AI to Pick Dividend Stocks And It Blew My Mind
Transcript
- 0:09Welcome to Roaming Returns, a podcast about generating
- 0:12a passive income with dividend stocks so you
- 0:15can secure your finances and liberate your life.
- 0:18What happens when you ask AI to pick dividend
- 0:20growth stocks and it accidentally makes the best
- 0:23argument for compounding? In this episode, we're
- 0:26running a little experiment. Can AI show the
- 0:28difference between low -yield dividend growers,
- 0:30mid -yield dividend stocks, and high -yield income
- 0:33plays over a 10 -year timeline? And results brought
- 0:36up something most investors overlook. The stock
- 0:39with the biggest yield today may not be the one
- 0:41that pays you the most later. We'll talk about
- 0:43yield on cost, dividend growth streaks, drip
- 0:46compounding, and the surprising sweet spot where
- 0:48boring dividend growers can start leaving high
- 0:50-yield stocks in the dust. So, hey, we're from
- 0:54the middle of nowhere here in... New Mexico,
- 0:57but this is going to be what we're doing going
- 1:00forward, I would think. So that's that. This
- 1:02week here, I wanted to talk about, I did a thought
- 1:05experiment with artificial intelligence last
- 1:07week. I thought for funsies that I would have,
- 1:11why not have AI pick dividend growth stocks to
- 1:13drive home the compounding point? So all I really
- 1:16did was I prompted Google Gemini with the following
- 1:19quote rules. I want a few low yield investments,
- 1:23less than 2%, a few mid. yield investments two
- 1:27to five percent and a few high yield investments
- 1:29more than five percent dividend yield with the
- 1:32desired purpose to show share price and dividend
- 1:35growth with dividend growth streak listed over
- 1:37a period of 10 years from a hypothetical ten
- 1:40thousand dollar starting base so that was literally
- 1:42all i typed in to the google gemini now i mean
- 1:45i have to caveat by because i use google gemini
- 1:49sometimes the gpt with or um is it gpt whatever,
- 1:53to chat with. So it kind of knows me, but I would
- 1:56think that you could have similar results if
- 1:58you didn't, if you just started Google Gemini
- 2:00on its own. From what it was saying to me that
- 2:03once it starts to learn you, then it starts to
- 2:06give out information that should appeal to you.
- 2:08So we'll get to that at the bottom. It's very
- 2:10interesting. What came out of the AI was interesting
- 2:13as AI included the 2016 share price, the 2016
- 2:16income for all the shares you accumulated, and
- 2:20the 2026 income. as well as the 2026 share count.
- 2:25And then they also provided a 10 -year track
- 2:27record with the yield on cost. What AI didn't
- 2:30include was current price and the current amount
- 2:32of the holdings. Like, so if you bought X holding
- 2:35in 2016, how many shares you currently have?
- 2:38It didn't include that. Full transparency, we
- 2:41did have to clean up the results as we kept putting
- 2:43the information in chart form that for some reason
- 2:46does not translate from Google Gemini into Google
- 2:49Documents. I don't know why. If they're both
- 2:51from Google, you'd think they would have like
- 2:52some compatibility thing going on there, but
- 2:54it was incompatible. Or you're just tech challenged
- 2:57because you didn't ask me about this. So I had
- 3:00to like have it relisted. I had to actually say,
- 3:02okay, that didn't work. I need you to relist
- 3:04it. So it came out in this list form. Because
- 3:07the AI introduced a concept outside of the desired
- 3:09output, I did have AI define what yield and cost
- 3:12is. Other than that, all the information here
- 3:14is listed as AI generated with actual minimal
- 3:17input for me. So I may have basically just got
- 3:20rid of my own job, but we'll see. Did you evaluate
- 3:22these, though, for your perspective? I did. They
- 3:25seem okay. Okay. That's what I wanted to know.
- 3:27Other than that, the... Other than that, all
- 3:30the information listed is AI generated. After
- 3:32we go over the information, I do have some observations
- 3:35about this data and this experiment as a whole.
- 3:37I just thought it'd be kind of a funsy thing
- 3:39to do because AI is like all people talk about
- 3:41and they're like, oh, AI is taking over jobs.
- 3:43AI is evil. We're in an AI bubble. If you use
- 3:46AI like it was kind of built to do, it actually
- 3:50is quite beneficial. We'll get to that at the
- 3:53end of this. Here we go. The complete 10 -year
- 3:56drip compounding breakdowns. The first one that
- 3:58it shot out was Texas Instruments, ticker TXN,
- 4:02and it came out with a 2016 share price of $60.
- 4:06With your hypothetical $10 ,000, you accumulated
- 4:09167 shares in 2016, which generated $253 per
- 4:14year. Isn't that who my TI -89 calculator from
- 4:17college high school was? Yeah. Texas Instruments?
- 4:20It was. Yeah. So if you compounded the shares
- 4:24over the last 10 years in 2026, you would have
- 4:28211 shares. You would actually have $1 ,200 per
- 4:33year in income. So the share count didn't go
- 4:36up very much. If you notice that, you went from
- 4:38167 to 211, which is at 33, 44. You only had
- 4:4344 different shares, but your income nearly quadrupled.
- 4:47It went from 253 to $1 ,200 per year. What it
- 4:51does, the 10 -year track record, 14 .1 dividend
- 4:55CAGR, which is basically dividend total return,
- 4:57and it currently has a 22 -year dividend payout
- 5:00streak. Your current yield today, if you bought
- 5:04it today, it would be 2 .5%. If you bought it
- 5:07in 2016, your yield on cost would be 12%. But
- 5:11that brings up a point we should probably discuss
- 5:13the yield and cost quickly. Yeah, because they
- 5:15need definitions to start. Okay, yield on cost
- 5:17is simply measuring how much cash a stock pays
- 5:20you today based on the original dollar amount
- 5:22you actually pulled out of your wallet to buy
- 5:24it, not what the stock is trading for in the
- 5:26news right now. So basically, if you bought the
- 5:29Texas Instrument, it probably wasn't at 2 .5
- 5:33% back in the day, but if you actually bought
- 5:35it back in the day, what you bought in 2016 is
- 5:3812 % now. Okay, so what he's saying is if you
- 5:42bought it for $10 back in 2016 and it is now
- 5:46worth $50 and the yield is gone to whatever based
- 5:50on the current $50, your personal yield, yield
- 5:53on cost, is not that 50 current yield that's
- 5:56going to be listed. It's actually going to be
- 5:57based on your original purchase price. Unless
- 6:00you actually have like share reinvestment compounding,
- 6:03which creates like a different cost basis, but
- 6:05it's essentially the gist. And that's the tricky
- 6:08part with these dividend aristocrats, dividend
- 6:11growers. If you open up any financial website
- 6:13or your brokerage, you're like a lot of like
- 6:15a lot of these things are going to be lower yields.
- 6:18Like, for example, Microsoft, if you bought Microsoft
- 6:20today, it would be 1%, probably less, a little
- 6:23under 1%. For those people who have owned Microsoft
- 6:25for the next number of years, that 1 % is meaningless
- 6:28because they bought it so long ago. And the dividend
- 6:30has been raised every single year. So their dividend,
- 6:33their yield on cost because all that is like
- 6:367%, 10%, or even 12 % depending on when they
- 6:38bought it. That's why you want the dividend growth
- 6:40stocks because they actually, when they raise
- 6:42it, say 5%, it actually starts to add up through
- 6:46time. So you can end up at 50%, even potentially
- 6:50100 % if these things are held long enough and
- 6:52passed down to like generations below. So that's
- 6:55where I think it gets a little, I don't know
- 6:57if you should say unfair, but like gaming the
- 7:01game. If you can actually inherit these things
- 7:03to your kids and your kids' kids and your kids'
- 7:05kids because the dividends that they're paying
- 7:06out are ridiculous based on cost yield. What
- 7:10I think the AI was trying to generate when I
- 7:13asked it this was if they bought Microsoft years
- 7:16ago and they got it for 1%, they got a dividend
- 7:18check for, say, $10. That same $10 now because
- 7:23the dividend increase would be like, say, $80.
- 7:27So then you take the $80 divided by the amount
- 7:30that you invest into it back in the day, not
- 7:31what you invest into it now. I think that's what
- 7:34I was trying to get at. It's not apples to apples,
- 7:37but it's the same concept of like that's how
- 7:40I evaluate stuff is like I basically whenever
- 7:42I get paid a dividend and I take it in cash,
- 7:45I actually subtract that cash amount from my
- 7:47total cost basis. I don't update the cost basis
- 7:51in the dollar form. But say, for example, whenever
- 7:55we get a check from NVDW, that $100 check comes
- 7:59out of what I initially put into NVDW. So my
- 8:02cost basis goes down $100. This is just, I think,
- 8:04the way AI generates that without doing the math,
- 8:08I think. But just think of yield and cost. Basically,
- 8:10it is the amount that your dividend check is
- 8:13worth today divided by your initial investment.
- 8:17how much money you actually put into it back
- 8:18in the day. Right. So all the examples on that
- 8:21are going to be based on 2016 data? Yeah. Okay.
- 8:24So that's number two. Did you finish number one?
- 8:28Yeah. So like if you look at Texas Instrument,
- 8:31for example, we have the 12 % ULINECAST. That
- 8:33means that $1 ,200 dividend check they got in
- 8:362026 divided by the $10 ,000 they initially put
- 8:39into it in 2016 gives you a 12%. So I think that's
- 8:43what they're doing. Okay. So that's that one.
- 8:46That's actually pretty sick that it grew that
- 8:47much. That is like a lot. That's $253 at $1 ,200
- 8:50because all we're really interested in more so
- 8:53than yield is how much we make. We want the cash.
- 8:56I don't give a shit if it yields 3%. I don't
- 8:59care. Number two that it spit out was United
- 9:06Health Group, ticker UNH. In 2016, the share
- 9:10price was $120. which bought you 83 shares in
- 9:142016. And your cash income back in 2016 was $208
- 9:18per year. Your current shares in 2026, if you
- 9:22dripped that from 2016 to now, would only be
- 9:2596 shares. You only accumulated 13 shares. But
- 9:29again, your 2026 cash income went up from $208
- 9:33to $888. That's insanity. That kind of ties into
- 9:37the next point, though, the 10 -year track record,
- 9:39which is 14 .0 % dividend CAGR with a 16 -year
- 9:42dividend payout growth streak. And your current
- 9:45yield in 2026 on this one is 1 .6%. but your
- 9:52yield on cost is 8 .88 because you're making
- 9:54$888 on $10 ,000 as opposed to $208. So that's,
- 9:59again, a lot of dividend growth for a very minimal
- 10:02yield. 1 .6 yield is kind of cheeks, but if you're
- 10:06making $888 after 10 years as opposed, that's
- 10:09$680 more. That shows some pretty good dividend
- 10:13growth, actually. The third one that it shot
- 10:16out that was a smaller one was the Home Depot
- 10:19HD. The price in 2016 was $125, which bought
- 10:26you 80 shares in 2016, and it generated you $221
- 10:29per year in 2016. Currently, you would have 98
- 10:34shares. Again, that's better than the two above
- 10:38it, though. That's 18 shares. And your cash income
- 10:41in 2026 would be $910 per year. Again, that's
- 10:45a huge, this is only 10 years. That's a huge
- 10:48step up from 221 to 910. And that has a 13 %
- 10:51dividend cager and it has a 17 year payout dividend
- 10:55growth streak. Your current yield today is 2
- 10:59.45%. Your yield on cost, if you take the 910
- 11:04divided by the 10 ,000, you're going to have
- 11:069 .1 % on what you actually put into it. Again,
- 11:10those are bonkers numbers. I know. That's why
- 11:13I told you about the inflection point on some
- 11:14of these. These are actually better than the
- 11:17THTA and stuff. That's what I was saying. There's
- 11:19a flip over point where these things become more
- 11:21lucrative in the long run. Okay. And then number
- 11:23four is the last of the low yielders, it looks
- 11:26like. Microsoft MSFT. In 2016, the price was
- 11:31$50 per share, which would have got you 200 shares.
- 11:34And your income in 2016 would have been $288.
- 11:38Current shares would be 223. So you only made
- 11:4123 shares in 10 years, but your cash went from
- 11:44288 up to 811. So that's not as good as the previous
- 11:48three, but it's still pretty damn good on a,
- 11:50like a low yield tech stock that everyone wants
- 11:52to get into. Right. Your 10 year track record
- 11:55is 9 .7 dividend cager. And Microsoft currently
- 11:59has a 21 year dividend growth streak. So your
- 12:02current yield today is 0 .85%. If you bought
- 12:05Microsoft today, it would be 0 .85%. Which is
- 12:08piddling. Not today, but like when I did this
- 12:10a couple days ago. And your new yield on cost
- 12:13is 8 .11%. So you're getting 8 .1 % on the $10
- 12:16,000 you invested in 2016. And it'll just go
- 12:20up from there? Bonkers. Now, this one is one
- 12:23of the mid ones. It was the remember the two
- 12:25to five percent to get this T -Row T -Row price.
- 12:28Oh, we so happen to have one of those. This is
- 12:31one that we talked about a while, six months,
- 12:34seven months, eight months ago around that time
- 12:36frame where I said that's so it's super undervalued.
- 12:38You should get into it. It has a pretty good
- 12:41track record with dividend growth streak and
- 12:43it was super undervalued at the time. Well, now
- 12:45this data here. If you hold it from 10 years
- 12:48from now, like you probably get comparable data,
- 12:51but not the exact same data. But in 2016, it
- 12:54cost $72 a share. And I think we actually got
- 12:56it cheaper than that in 2026. But I digress on
- 12:59that point. You would have bought 139 shares
- 13:03in 2016 and your cash would have been $300 per
- 13:06year in 2016. You currently, if you held it for
- 13:09the 10 years and had the drip on, you would have
- 13:11192 shares. So that's a pretty big share increase
- 13:15compared to the four above it. The 2026 cash
- 13:19income would be 965. So that's like more than,
- 13:22what is that, 66%, 67 % growth in 10 years. That's
- 13:27not great, but it's not as good as the other
- 13:29ones, but it's still pretty good. 67, that's
- 13:32like over 100, 300%. The 10 -year track record
- 13:36on this one is only 8 .9 % dividend cager, but
- 13:39this one does have a 39 -year dividend growth
- 13:41streak. So that means the dividend growth is
- 13:44not near as much as the other ones. I don't even
- 13:46have to look at it. I can tell you that the four
- 13:48that we mentioned have a higher growth rate than
- 13:50TRO does. Your current yield today in 2026 is
- 13:544 .65, but your yield on cost would be 9 .65
- 13:59if you held it for the last 10 years. So it's
- 14:00like double, 100%. 12 % around there. It's still
- 14:05freaking good. Yeah. And this is the first one
- 14:10that's in the high yielders. This is everybody's
- 14:13favorite tobacco stock, Altria. And by that we
- 14:16mean Tim. MO. In 2016, you could have got MO
- 14:20stock for $60. In 2026, you could get MO stock
- 14:24for like $60. So the price hasn't really went
- 14:26too far. But in 2016, you would have got 167
- 14:29shares. which would have generated you $406 per
- 14:33year in 2016. In 2026, if you had the drip on,
- 14:37you would have almost 300, 297 shares. So that
- 14:40one, you got a lot of shares. That's like the,
- 14:45I think that might be the most on the list. Your
- 14:47current income in 2026 would be 1 ,210. per year,
- 14:52which is bonkers on a $60 stock back in the day.
- 14:56This one has a 10 -year track record of 5 .3
- 15:00dividend cagers. So that one, again, like minimal
- 15:02growth is what it looks like because that's crazy.
- 15:04The other ones that have like 14, 12, 11, and
- 15:07this one has 5 .3%, but it does have a 54 -year
- 15:11dividend growth streak. And your current yield
- 15:14today, if you bought Alturia today, it would
- 15:16be 8 .3%. I think it's not that high. I think
- 15:18it's like 7 .8 % now, but it's around there.
- 15:21Your yield on cost would be 12 .10%. So your
- 15:25$10 ,000 in 2016 is now generating 12 % as opposed
- 15:29to the 8 .3%. So that's fun. Num -num. Speaking
- 15:34of, time for a num -num. The number seven here
- 15:37is Snap -on Incorporated SNA. Jesus. This is
- 15:42another one of those. This is one of the middle
- 15:44ones is 2 % to 5 % range. In 2016, the share
- 15:49price for SNA was 155, which it would have only
- 15:52got you 65 shares, which would have got you $164
- 15:56per year is all. Currently, if you had to drip
- 16:00on the entire 10 years, you would have 82 shares.
- 16:03So you only went up 18 shares. But because of
- 16:06dividend growth, you're... 2026 cash income is
- 16:10now $800 per year. That means your 10 year track
- 16:13record is 14 .4 % dividend CAGR. And this one
- 16:17only has a 16 year dividend growth streak. Your
- 16:21current yield today is 2 .55%, but your yield
- 16:24on cost is 7 .99%. So this one did really well.
- 16:28And like, what the hell is Snap -on? Is that
- 16:29like the... Those are those tools. Tools? That
- 16:31every mechanic buys and like ends up with loans
- 16:33and payouts and they end up in debt forever because
- 16:37they have to buy the Snap -on brand. Because
- 16:39that's like just like an... Cursor review on
- 16:41this one, like your shares didn't go up that
- 16:44much. It was like, what, 65 to 82, 17 shares
- 16:47difference. But your money went up like almost
- 16:49four times. You almost made 400 % on your dividend
- 16:52income. So this one had to have some wicked growth
- 16:54in the last 10 years in the dividend department.
- 16:57Maybe not the price. I don't know. Number eight
- 17:00is the second high yielder or no, not now. It
- 17:04actually, I think it's still my, this one's on
- 17:06the line. It could be high yielder because remember
- 17:08the framework is 5 % plus and then two to five
- 17:11with the mid. This one, I think it might be 6%.
- 17:14I don't know. UPS. This is one that we own as
- 17:18well. I really think this one's supremely undervalued.
- 17:21And I'll tell you why here in a second, because
- 17:23in 2016, it costs $105. And in 2026, it costs
- 17:26like $92, $93. So this one actually had price
- 17:30depreciation. But in 2016, your $10 ,000 would
- 17:34have got you 95 .2 shares, and that would have
- 17:37generated you $297 per year in income. Currently,
- 17:42if you had the drip on for the 10 years, you
- 17:44would have 129 shares. Again, I got a lot of
- 17:46share growth in this one, but you didn't see
- 17:49it. It didn't equate to the one with number seven
- 17:52where the money went up like 400%. This one,
- 17:54your money, you're only making $838 per year
- 17:59in UPS, so it only went up. a little over 160%,
- 18:03maybe, around there. 10 -year track record, again,
- 18:07you'll see the 7 .7 % dividend cashier. So that's
- 18:12not anywhere, but that's like exactly, it's almost
- 18:14half. The snap -on was 14 .4, and this one would
- 18:18be 15 .4 if it was 100%. So that's kind of ass.
- 18:22And UPS only has a 16 -year dividend growth streak
- 18:24going on currently. Interesting. But the current
- 18:27yield, as I said, when I wrote this, it was 4
- 18:28.75. Again, I think it's above 5 % now. But your
- 18:32new yield on cost would be 8 .38%. So, like,
- 18:35this one, you barely doubled your yield in, like,
- 18:38the last 10 years. So, it's kind of been a shit
- 18:4010 years. Okay, here's another mid one that I've
- 18:44been, like, I looked and put on the watch list.
- 18:47Forever ago? Forever ago. I don't know if I'll
- 18:49ever get into it, though. It's just, it's really
- 18:50pricey. It's Lockheed Martin LMT. In 2016, it
- 18:56was $220, and I think it's like $500 and some
- 18:58dollars now. It's a crazy amount. So your $10
- 19:01,000 would have bought you 45 shares in 2016,
- 19:04which would generate you a pretty good amount
- 19:06of money, though. $300 per year on 45 shares
- 19:08is pretty good. Currently, if you had the drip
- 19:11on for the last 10 years, you'd only have 59
- 19:13shares, so that only went up 14 shares maybe.
- 19:17It's still not terrible considering that price.
- 19:19But your income would have increased from $300
- 19:21per year in 2016 to $737. per year in 2026, so
- 19:26that's more than 100%, so that's good. The 10
- 19:29-year track record, this one only had a 6 .7
- 19:31% dividend cager, which means they had really
- 19:33low payouts. We're talking 1 % dividend increases,
- 19:38but it does have a 23 -year dividend growth streak.
- 19:41Current yield today is 2 .40%. Your yield on
- 19:44cost from 10 years ago would be 7 .37%, so that
- 19:47one is much better than UPS, much better from
- 19:50that perspective. But UPS, you can get two shares.
- 19:53You get like two to one. And the last one for
- 19:56funsies, because it gave me a list of like 15,
- 20:00but like I just put this one on the list, Costco.
- 20:04And this one, I've been thinking about getting
- 20:06in this one for a long time. It's like $1 ,000
- 20:08a share because I know they're probably going
- 20:09to do a stock split soon. Oh, so hold off and
- 20:12wait until it does one? Yeah, so I'm probably
- 20:14going to wait until they do another stock split.
- 20:15Anyway, Costco in 2016 was $150 per share. And
- 20:20that would have got you 67 shares in 2016. That
- 20:23would only generate you $120 per year in cash.
- 20:26Current shares will only be 73. So you only gain
- 20:29seven shares in 10 years. But your income, damn
- 20:33near, almost 200 % dividend income growth. Your
- 20:382026 income would be 340 per year. So it's doing
- 20:429 .9 % dividend cagers. So again, anything under
- 20:4510 % I'm starting to see is like their increases
- 20:47are like 1%. Less than 1%, maybe 1 .3%, whatever.
- 20:52But it has a 22 -year dividend growth streak.
- 20:54And current yield today is 0 .55%. But this is
- 20:57the one that really drives home the point, like
- 21:01holding them for 10 years. It's not going to
- 21:04look like much, but your yield on cost is 3 .4%.
- 21:08So that is like almost 700 % growth in your yield,
- 21:12right? 518 point. Like 2%. So it's 518 % growth.
- 21:18Yeah. So that's crazy. So this is one of those
- 21:22times where you have to kind of throw out the
- 21:24yield. And that was kind of the whole thought
- 21:27experiment. Like the yield's not as important
- 21:29as the dividend growth. And we're probably going
- 21:32to focus on that more because you've actually
- 21:34seen the main portfolio is kind of deviating
- 21:37from the high yielders into the dividend growers.
- 21:40And there's a reason for that is because that
- 21:41main portfolio is going to have to supplement
- 21:43our life. for years and years and years. So we
- 21:46kind of need that to be growing as opposed to
- 21:48just generating cash. We're still going to have
- 21:50the cash ones in there because I'm going to be
- 21:51popping that into either the bank accounts or
- 21:53into other stuff. Well, plus it always allows
- 21:57you to take advantage of one of these things
- 21:58when they drop down, like you said, UPS and I
- 22:02forget which other one was down below their 2016
- 22:04price. T -Row. T -Row. And we picked up both
- 22:08those on the cheap. We picked up both UPS when
- 22:10it was under like $90. We picked up T -Row when
- 22:12it was under, I think it was under $60. It might
- 22:15have been under $50. I forget exactly. And I
- 22:17would not put it past the people who actually
- 22:18owned these things back in 2016 that they would
- 22:21have bought the dip as well if it would have
- 22:22come down because that would have just dropped
- 22:24their cost bases down ridiculously. So that's
- 22:26two lessons. The first lesson is have money in
- 22:28dry powder. The dry powder needs to be growing.
- 22:30That's why we have the THTA and that's why we
- 22:32have all that stuff. So that we can tap in, like
- 22:35I can actually pull money out of the dry power
- 22:36when I need to actually put it into something
- 22:38that I see has a crazy dividend growth. I don't
- 22:43care about the dividend streak as much as the
- 22:44dividend growth. If you're looking like 9%, 10%,
- 22:4711%, 12%, 13 % in that range, and that's their
- 22:49growth for the last 10 years, and they go on
- 22:51sale, you need to pick those up. Right. And you
- 22:53need money to do that. So it's good to have liquidity.
- 22:56So a couple of interesting things I got from
- 23:00this experiment. was the first is the AI kept
- 23:04talking a lot about a company only has a finite
- 23:07amount of net income. And what the AI generated
- 23:10was it generally splits it into three buckets.
- 23:13So they either reinvest it in the business to
- 23:16grow, they either buy back their shares to lower
- 23:19the amount of shares on the market, which thus
- 23:21driving the price up, or they pay it out as dividend.
- 23:23That's the three buckets that generally the net
- 23:25income from a company goes into. Which is true,
- 23:28but I never thought of it that way. I didn't
- 23:31either. But that makes a lot of sense from a
- 23:33dividend grower or the dividend companies. Because
- 23:35that's one of the things we may touch on in the
- 23:38near future is if you can find companies. I know
- 23:40we talked about it before, but you can find companies
- 23:42that regularly buy shares back. That is just
- 23:46as good as a 5 % or 10 % dividend growth from
- 23:51year to year because taking those shares off
- 23:53the market drives the price up. So your initial
- 23:57shares are going to be worth a lot more. And
- 24:00they have to pay out less in dividends. So it
- 24:02actually, in the long run, actually helps with
- 24:04their net income. So it's interesting. But one
- 24:07of the things that way, like once this all came
- 24:09out, I started talking with AI and one of the
- 24:101 % to 2 % boring growers like Microsoft, Visa,
- 24:14UnitedHealth, they have a ridiculously low payout
- 24:17ratio, obviously, because they're only paying
- 24:19like 1 % of their net income into dividends.
- 24:23And that generally it's between 10 and 30%. So
- 24:25that means they can keep 70 to 90 % of their
- 24:27cash to build data centers, to fund research
- 24:29and development, or buy competing companies.
- 24:32Because they actually aggressively reinvest,
- 24:35their earnings explode. When earnings explode,
- 24:37they can easily raise their dividends by 12 to
- 24:3915 or 20 % a year without sweating. So like the
- 24:411 % to 2 % grower, like the boring growers, have
- 24:45their place in your portfolio. Whereas the 5
- 24:48% high yielders like Altaria, Verizon, REITs,
- 24:52these companies are usually very mature, which
- 24:54means they have a slower growing business and
- 24:56they already have a high payout ratio of 75 to
- 24:5990 % in that range. And it doesn't even apply
- 25:01to the ones that are like 8%, 7%. You've seen
- 25:04some of this like T -Row price at 4 .5%. You
- 25:07see that it has a higher payout ratio and it's
- 25:11a slower growing one. So it's kind of a slow
- 25:13burn, I guess would be the way to put it. Because
- 25:17they're basically giving almost all their cash
- 25:18to shareholders on day one, they have very little
- 25:20money left over to reinvest in growth. As a result,
- 25:22their business growth slows to a crawl, meaning
- 25:26their dividend increase are structurally limited
- 25:28to matching inflation 2 % to 5 % per year. So
- 25:31that is an important lesson to take from that
- 25:35is the 5 % yielders have less money to actually
- 25:38research and develop to buy competing companies
- 25:41because they're giving it all to the shareholders
- 25:43to begin with. Right. So longevity of these might
- 25:46actually be questionable is what I'm hearing.
- 25:47Yes. It's possible. Or sustainability of continued
- 25:50increase. It'll probably increase and then plateau
- 25:52or something. These are the ones that you see
- 25:54like when the dividend news comes out and they'll
- 25:57be like, oh, such and such has their 41st year
- 26:01of dividend increases. It's the ones that do
- 26:02like one penny. So technically they have a dividend
- 26:06growth. dividend growth streak intact because
- 26:09they did raise their dividend year over year,
- 26:11but it's only one penny or two pennies. These
- 26:14are the ones that do that when they have a high
- 26:17payout ratio and they have all their money being
- 26:20dispersed to shareholders. So it sounds like
- 26:22they bait you in the beginning with higher payouts
- 26:25and then they... So the AI basically compared
- 26:29this to the Taurus and the Hare of, what is that,
- 26:31Fable from back in the day. The hare, they start
- 26:34at 1 .5%, but they grow their payout like at
- 26:3715 % per year. By year 10, your yield on cost
- 26:40climbs to roughly 6 .1%. By year 15, it hits
- 26:4412 .2%. which is not it's just bonkers nuts whereas
- 26:49you then you have the hortus they started the
- 26:51hortus the hortus the hortus the tortoise which
- 26:53starts at like a six percent yield but only grows
- 26:56their payout ratio three percent by year by year
- 26:5810 your yield and cost is eight percent which
- 27:01is just a little bit higher than the hair but
- 27:05then by year 15 it's just 9 .3 percent so by
- 27:08like a year 10 like the 10 to 12 year mark the
- 27:11boring low -yielding stocks actually cross the
- 27:13line where they're paying you more on your money
- 27:15over the long term than the fucking ones that
- 27:17you got that you thought would be oh i'm even
- 27:19making so much money from these yeah interesting
- 27:21and after 10 to 12 it actually only widens further
- 27:26every year for that because the the hair ones
- 27:28are going to keep growing their dividend 10 11
- 27:3012 13 and the tortoise ones are only going to
- 27:33be growing their dividend one two three percent
- 27:35so like that's just going to it's just going
- 27:36to widen I feel like that analogy should have
- 27:39been flipped the other way around. No, because
- 27:41the hare goes faster. Yeah, they do go faster.
- 27:43So the hare is like the 1 .5 low yielders. They
- 27:46go way faster than the tortoise ones, which are
- 27:48slow and steady. No, but it's the appearance
- 27:50of the slow and the hare is the one that dies
- 27:53at the end of the race and the tortoise wins.
- 27:55You have that analogy backwards. Where did the
- 27:57AI do that? The AI did this. That's awesome.
- 27:59I feel like that's backwards. I love when AI
- 28:02does weird shit. So what it means for people
- 28:04is if you need cash right now, don't call JD
- 28:07Wentworth. Oh, that was good. If you're retiring,
- 28:14you actually need the high 5 % starting yield.
- 28:17You don't have a 10 -year runway to wait for
- 28:20Microsoft's 1 % yield to compound into something
- 28:22ridiculous. you trade future growth for immediate
- 28:25cash flow for survival. So that's kind of like,
- 28:27if that's where you are, you should be looking
- 28:29at the 5 % years. Whereas if you have more than
- 28:3410 % or 10 years plus runway, you kind of want
- 28:37to have a majority of your stocks in the wealth
- 28:40building node where you're buying. We're buying
- 28:43a 5 % stock with a 2 % growth is a mathematical
- 28:45mistake. You are actually much better off buying
- 28:47the boring 2 % yields with a double digit growth
- 28:49rate because the drip engine and dividend hikes
- 28:51will leave the high yielder in the dust. So like
- 28:54if you have more than 10 years, you kind of want
- 28:56to have a majority of your portfolio in the ones
- 28:58that are only yielding 1, 2, 3%. It totally has
- 29:01an analogy backwards. But this also would come
- 29:04into play too that you may need to reconsider
- 29:06if you're planning to leave your stocks as a
- 29:07legacy. You probably also want to get in. Even
- 29:10if it's not for you, you want to get into the
- 29:12ones he was just talking about, the ones that
- 29:14have the lower. So then once we saw this, because
- 29:18I've had this theory for some time and I actually
- 29:20went with the AI to like either confirm it or
- 29:22tell me I'm full of shit. I think what you should
- 29:25be looking for, no matter what investor you are,
- 29:28is the ones that have a 10 to 20 year dividend
- 29:29growth streak. going on currently. I understand
- 29:32the dividend kings get all the publicity because
- 29:34it's like 50 years. And I understand like the
- 29:37aristocrats get a lot of public because they're
- 29:38over 25 years, but hear me out. The ones, the
- 29:4210 to 20 range is where the sweet spot is. And
- 29:44I'll go actually go into this because the AI
- 29:46actually generated some, some confirmation to
- 29:48this theory. You have phase one, which is the
- 29:51incubate, the incubation phase, which is the
- 29:53zero to nine years. dividend growth streak. The
- 29:56company just started paying the dividend. They're
- 29:58growing like crazy, but management is still incredibly
- 30:00paranoid because they actually haven't survived
- 30:03any of the nasty, volatile things. The growth
- 30:06reality, this is where the highest mathematical
- 30:08cager often lives, frequently 15 % to 25 % plus,
- 30:12but it also is super highly volatile. Because
- 30:15the streak is short, you have no historical proof
- 30:17that management will protect the dividend during
- 30:19a recession, a downturn, whatever, a pandemic.
- 30:22They haven't had to prove any of that. They just
- 30:25basically have been growing in the sideways or
- 30:26bull market. So you really have no historical
- 30:28proof. They very well might work, but you kind
- 30:31of want to reduce your risk. You kind of want
- 30:33to have more mathematical certainty. Screeners
- 30:37seldom capture this as well. So if you use a
- 30:39screener like in Schwab, we have Schwab. If you
- 30:41use the dividend screener in Schwab, they seldom
- 30:44capture this. A five -year streak often looks
- 30:47like a temporary blip, and many of these companies
- 30:49have microscopic starting yields, think 0 .5%.
- 30:51Meaning even with massive growth, it takes too
- 30:54long to generate meaningful cash. So that's a
- 30:57zero to nine -year dividend growth streak range.
- 30:59It's very tricky because a lot of the companies
- 31:02haven't had to face any... financial problem
- 31:05so they have never had to cut the dividend and
- 31:07the 15 to 25 growth rate is it sustainable we
- 31:12don't know because we've never seen any downturns
- 31:14couple that with like the screeners don't capture
- 31:18it like they did it's very difficult to find
- 31:20like any stability metrics in a company that
- 31:23has only been around that long whereas if you
- 31:26do the phase two which the ai labeled the sweet
- 31:28spot It was the 10 to 20 -year dividend growth
- 31:31streak range. The business model is fully scaled,
- 31:34wildly profitable, and throwing off way more
- 31:35cash than it can internally reinvest. Management
- 31:39at this point is committed to a dividend -centric
- 31:42corporate culture. So the growth reality, this
- 31:45is where you find the golden intersection of
- 31:47high starting yields, 2 % to 4 % range, mixed
- 31:50with double -digit cagers, dividend growth, 10
- 31:53% to 15%. The math works in this case because
- 31:56the company has survived at least one full economic
- 31:58cycle, like 2020, and actually had a downturn
- 32:01and defended its payout, provided you proved
- 32:03for its safety and kept growing it. Yet the business
- 32:07isn't so massive that its growth has stalled.
- 32:10That's the key word, stalled. They still have
- 32:12room to expand their profit margins and market
- 32:13share, which funds those massive annual rages.
- 32:16Rages. Rages. Rages. Rages. For us, phase three
- 32:21is the elite aristocrat phase, the 25 -year plus
- 32:24dividend growth streak. The business at this
- 32:26point is a massive lumbering giant. Think consumer
- 32:28staples, utilities, or legacy industry, industrials.
- 32:32The growth reality in this phase is the average
- 32:3510 -year dividend occasion for the S &P 500 dividend
- 32:38aristocrats typically hovers around 5 % to 6%.
- 32:41So you're getting like half the dividend growth.
- 32:45Why the growth slows is when a company has raised
- 32:47its dividend for 30, 40, or 50 years, their payout
- 32:49ratio is usually pushed to its absolute limit.
- 32:52They are already paying out 70 % to 80 % of their
- 32:54earnings. Because they are mature, their net
- 32:56income is only growing at 3 % to 4 % per year.
- 32:59And to protect their legendary streak without
- 33:01running out of cash, management is forced to
- 33:02slow dividend raises down to token inflation
- 33:04matching increases, often just a few pennies
- 33:07a year. So I've seen that. The dividend aristocrats,
- 33:12they don't grow very fast. And the newbies grow
- 33:16fast, but there's like no certainty that they'll
- 33:19keep growing. So that's kind of why I settled
- 33:20on the 10 to 20 years seems to be the sweet spot.
- 33:23Well, doesn't one of these kind of turn into
- 33:25each one of these categories at some point? It
- 33:28does. So that's why I like the 10 to 20 years
- 33:30because it's the zero to nine will tick over
- 33:32into like the golden years. 10 to 20 where you
- 33:35can make the most money. Right. Because the first
- 33:37phase is more risk because they haven't been
- 33:39able to determine they can sustain intelligently
- 33:41through a negative market situation. Then after
- 33:44they do the 10 to 20 years, they'll actually
- 33:46eventually turn into 25 plus years. And at that
- 33:48point, everything will slow down because they've
- 33:50already bought up all their competitors and they've
- 33:52already did the research and development. At
- 33:54that point, they're just maintaining shelf space
- 33:57with their products. I don't know. So why the
- 34:00AI thinks the 10 to 20... Your dividend growth
- 34:03streak focus wins is by systemically targeting
- 34:06companies with 10 to 20 year streaks. You are
- 34:08intentionally avoiding the two biggest traps
- 34:10in dividend investing. You're avoiding the unproven
- 34:14rookies, zero to nine years. Oh, like I just
- 34:16said. That might cut their dividend the moment
- 34:18a micro headwind hits. And you're avoiding the
- 34:20exhausted old timers, the 25 plus years whose
- 34:22days of hyper compounding your yield and cost
- 34:25are mathematically over. So you're looking for
- 34:27like. companies in their prime, which is if you
- 34:30were like, if you're going to relate it to people,
- 34:32like we have some stuff in Kings, but we bought
- 34:34them at like massively discount prices. Did we
- 34:36not? Our yield and cost is going to be really
- 34:38good. And that was because of bottom. As for
- 34:40this experiment in the hole, on the hole, in
- 34:43the hole, is it not in the hole, in the hole,
- 34:45on the hole? No, no. In the hole is bad. Take
- 34:48that out. I think people are using AI incorrectly.
- 34:53The results you get are directly correlated to
- 34:55the input that you're putting into it, the input
- 34:57that you offer. If you ask vague questions, you're
- 34:59going to get vague response or common responses.
- 35:03The biggest thing I've noticed with a bunch of
- 35:04my friends that have asked me, how the hell are
- 35:06you getting inputs that you're getting? I don't
- 35:07treat it like a freaking oracle. I don't give
- 35:09it vague shit to see if it can read my mind.
- 35:12I tell it what I'm trying to freaking do. I give
- 35:15it bounds. I give it scope. I give it perspective
- 35:17and whatnot. I tell it when I'm a newbie and
- 35:20I need things broken down into idiots step by
- 35:22step. And then I challenge it. Well, like then,
- 35:25so if you went back to what the original inputs
- 35:27for this were like kind of vague and they gave
- 35:29me like kind of like vague stuff, but because
- 35:31it's like interacting with me a couple of times
- 35:32before it knows that I'm doing dividend investing.
- 35:35So it actually. I generate stuff and we'll get
- 35:37to that in a minute. It also knows that he likes
- 35:38to go deeper and not just surface level stuff.
- 35:40We'll get into that in a minute. But at the same
- 35:42time, if I was going to do it, like if I was
- 35:44a newbie doing this and I went into AI, I would
- 35:46literally ask every data point that I was interested
- 35:48in, like the yield on cost, dividend growth rates
- 35:52for the one year, the three year, the five year,
- 35:53the 10 year. I would ask it for share prices
- 35:56over the years, like has the price appreciated,
- 35:58depreciated over 10 years, over 20 years, whatever.
- 36:02So then. Because I remember I said at the top
- 36:05that it actually put information in that I didn't
- 36:07specifically ask for. And I actually did have
- 36:09a conversation with AI after the initial data
- 36:11was produced. And I asked it why it offered data
- 36:14I didn't ask for. The AI responded with, when
- 36:19AI provides information you do not explicitly
- 36:21request, like yield on cost or specific 10 -year
- 36:23performance data in your example, it is generally
- 36:26due to three primary drivers. Predictive patterns,
- 36:28helpfulness -based tuning, and objective context.
- 36:31These are interesting. So the first one is the
- 36:34helpfulness heuristic. I like the quads. Yeah,
- 36:38the most modern AI models are trained using a
- 36:40process called reinforcement learning from human
- 36:43feedback. During this training, human raters
- 36:46evaluate model responses. And then consistently,
- 36:49models that provide comprehensive extra value
- 36:52are rated higher than those that provide maliciously
- 36:54compliant answers doing only exactly what was
- 36:57asked, even if the result is useless. So they
- 36:59actually have humans that actually go and say,
- 37:01well, you did this, you're good. So basically,
- 37:03it's like a star system. Yeah. Because the model
- 37:06has been reinforced to be an assistant, it is
- 37:09programmed to anticipate follow -up questions.
- 37:11If you ask for a list of dividend stocks, the
- 37:12model thinks users who ask about this almost
- 37:15always want to see the performance metrics to
- 37:17make a decision. By including that data, it attempts
- 37:19to save you time of having to ask the next logical
- 37:22question. Yeah. So that's segment one. They want
- 37:24to be helpful because they want you to keep using
- 37:26them. And they want to get their stars. So the
- 37:30second thing that it brought up was pattern matching
- 37:32and training data. LLMs operate on probabilistic
- 37:36prediction. They are trained on vast amounts
- 37:39of human discourse, financial blogs, analyst
- 37:41reports, textbooks, and forums. And in those
- 37:44data sets, whenever a list of dividend stocks
- 37:46is presented, it is almost... Never presented
- 37:48in isolation, the model is essentially completing
- 37:51a pattern. When it sees the prompt, list companies
- 37:53for dividend growth, the statistical completion
- 37:55of that request, and the model's training data
- 37:58involves columns for ticker, name, current yield,
- 38:01dividend growth rate, 10 -year total return,
- 38:03yield on cost. To the model, providing just the
- 38:07names would be like a human writing a sentence
- 38:09and stopping halfway through it. It's simply
- 38:10finishing the thought as they exist in the training
- 38:13data. Which is very interesting. Like so you
- 38:16could ask like partial questions and it'll give
- 38:18you like full blown responses. It does if there's
- 38:21a pattern for it, though. But again, if you give
- 38:24it an incomplete ask, you might not get the complete
- 38:28output. I think humans tend to not realize what
- 38:31they're asking all the time. And AI sometimes
- 38:34can intuitively leap through these patterns,
- 38:36but not always. And that's the part you got to
- 38:38be careful about. Like he was making fun of me
- 38:40because I like write books to mine. She does.
- 38:42She writes more. She has said more to AI than
- 38:45she said to me in like 12 or 13 years. Just in
- 38:47a prompt. I wanted to know how I think. I wanted
- 38:50to know what I'm looking for. I wanted to see
- 38:51patterns in my actual thought process. But that's
- 38:54me. And then the third thing that they said that,
- 38:56remember, it said there was three things it does.
- 38:57The third one is objective contextual needs.
- 39:00The model is designed to optimize for utility.
- 39:02In financial context, a list of company names
- 39:04is objectively less useful than a list of companies
- 39:07with supporting metrics. The model infers that
- 39:10you are conducting an analysis and therefore
- 39:11it adds data points that are necessary for the
- 39:13analysis to be meaningful. Garbage in, garbage
- 39:17out. The garbage in, garbage out. Remember I
- 39:19said, like, if you're vague with it, it's just
- 39:21vague back. There is a distinction between prompting
- 39:24for information and prompting for a specific
- 39:26output format. A vague prompt would be, tell
- 39:28me about dividend investing. The AI will dump
- 39:30a massive amount of unrequested data to cover
- 39:32all bases. If you do a precise prompt, list three
- 39:35companies for each of these categories, low yield,
- 39:37mid yield, and high yield. Provide only company
- 39:39names and tickers. Do not include performance
- 39:41metrics or yield on cost. When a user gets extra
- 39:44information, it is often because they failed
- 39:46to place constraints on the AI. Users often mistake
- 39:49an AI for a human assistant who can read their
- 39:51mind. In reality, the AI is a logic engine that
- 39:54needs strict negative constraints do not include
- 39:57to narrow its output. It also needs scope to
- 40:01or perspective, I guess, would probably be the
- 40:03better one. Basically, the more information you
- 40:06can give it on what the hell you're actually
- 40:07looking for, the better results are going to
- 40:08get. That's one of the things I got from this
- 40:10experiment was like, I don't think people need
- 40:13to actually subscribe to the people that have
- 40:16like the AI. algorithms to actually trade stocks
- 40:20you can create your own algorithm right and you
- 40:22don't even need to do anything crazy you can
- 40:24actually just talk to your freaking ai get it
- 40:26to know you because your better investment is
- 40:29something that's suited to you anyway your metrics
- 40:33your time horizon your amount of money your specific
- 40:36goals So if you feed all that into it, there's
- 40:40such a dynamic thing. It's so much better than
- 40:42just a normal Google search from before where
- 40:44you literally had to pull every piece out, synthesize
- 40:46it, take forever, dig deeper, go look for weird
- 40:48outliers. Because if you're not the same situation
- 40:50as somebody else, you have to dig, you have to
- 40:52look at reviews and you have to do whatever.
- 40:53And you have to like piece pieces together that
- 40:55might not fit. The freaking AI tools now are
- 40:57so amazing because they take stuff that used
- 40:59to take me freaking days to figure out in like
- 41:0330 seconds. The takeaway from the investing front
- 41:07is you legitimately could use your AI to generate
- 41:11a watch list and evaluation metrics so that you
- 41:16can evaluate stocks before you put them on the
- 41:18watch list. And then you could actually program
- 41:20your AI to basically not alert you, but just
- 41:24go in and say you have a watch list of 75 stocks
- 41:27and you have an evaluation metrics that the AI
- 41:29helped you with. You can go in once a week and
- 41:31be like, so any of my 75 stocks undervalued?
- 41:34It'll go through and it'll look at all the points
- 41:36that you put together to actually evaluate the
- 41:39stock. And it'll say, yeah, these three are undervalued.
- 41:41So this will save you so many buttloads of time.
- 41:45And a lot of money, especially if you're paying
- 41:47for AI bot trading or whatever the hell they
- 41:49have now. Those are just generalized. I don't
- 41:53think they'd be as specific if you took this
- 41:54extra step and did this yourself. You could potentially
- 41:56get rid of your financial advisor. I don't know
- 41:58how much those people are paying for fees. 1%,
- 42:002%. That adds up over time. And you could totally
- 42:03be doing this yourself. Literally, I could tell
- 42:05you right now I have an equation in the works
- 42:07that I'm working with AI about with ETFs and
- 42:10high -yield ETFs specifically to determine a
- 42:12couple formulas where we can actually extract
- 42:15if a high -yielder, based on historical data
- 42:18and current data, has what the probability is
- 42:21in the future, what the NAV decline, NAV erosion
- 42:24is going to be in the future. I also have...
- 42:27constructed a actual valuation equation using
- 42:30ai you ai's help where i can evaluate all the
- 42:34different ones that pop up on my radar through
- 42:36my readings like this this stock looks undervalued
- 42:39so we plug it into the actual valuation equation
- 42:41and it'll and it'll come up this is overvalued
- 42:44this is undervalued so like i've used ai to vastly
- 42:46reduce the amount of time i have to devote each
- 42:50week to Just data collecting. Stock research
- 42:52because it collects all your little, it'll like
- 42:54scrub the internet and get all the information
- 42:56I need. Help me create a formula. Once I have
- 42:59the formula, I can literally just say this, it's
- 43:01Friday. What's going on? Like is anything, anything
- 43:03in the, in the buy range? Yeah, it's pretty sweet.
- 43:06And that's literally just the tip of the iceberg.
- 43:09I'm assuming down the road, I'm going to have
- 43:10more crazy shit going on with AI. Right. But
- 43:14like the experiment itself was literally just
- 43:16to dump in like vague stuff and see what AI generated.
- 43:18AI actually did a pretty good job. like it actually
- 43:22got like a lot a lot of output for me to go through
- 43:25and like oh i never thought of that like i never
- 43:27thought of yield on costs right and that's what
- 43:29i'm saying like when i talk to it i ask it a
- 43:32question and because it's been talking to me
- 43:34so long or i actually use chat gpt and i program
- 43:36that i actually program ais to have specific
- 43:39perspectives and stuff so i'll actually run it
- 43:41through different ones to see if i get different
- 43:43lenses and one of them is actually built to like
- 43:45systems think and filling gaps and another one
- 43:48is actually built to challenge and to not assume
- 43:50and that assume assume growth not complacency
- 43:52and like it's just all this crap but what what
- 43:54I do is I'm very complex I'm very synthesized
- 43:57like synthesizing and sometimes you just don't
- 43:59know to ask certain questions because you just
- 44:01don't know terminology you're not an expert you're
- 44:04not I don't know, versed in the topic or something
- 44:07like sometimes the clarity actually comes from
- 44:10slightly like weird edge stuff. Like he was just
- 44:12saying, like, those are things you wouldn't have
- 44:14even known to look at. Like we've been doing
- 44:15this how long? And you never even thought to
- 44:17incorporate that in because it just wasn't something
- 44:18that. Well, it's because of something like I
- 44:19have a spreadsheet that I automatically do. I
- 44:22never actually thought to extrapolate that out
- 44:24to stocks we don't own. And even just some of
- 44:27these other like data points. So it's. But because
- 44:30the reason I didn't think to extrapolate it out
- 44:31to the stocks we don't own is because it is kind
- 44:34of tedious and time consuming to keep track of
- 44:36like 100 stocks to begin with. So like I don't
- 44:38have the free time to be like, oh, I'm going
- 44:40to do these 100 and see what's going on for like
- 44:41do a back test and stuff like that. And the AI
- 44:43literally can just do it in three seconds. And
- 44:45I think one of the reasons that people don't
- 44:48invest or may be hesitant or haven't started
- 44:51or one of the fears they have is that it's going
- 44:54to take them a lot of time. So it's like use
- 44:56AI to cut that time down so you can get started.
- 44:59If that's your hangup, use this as a tool. That's
- 45:01literally what it's here for. Yeah, it's not
- 45:02like AI is more than chatbot. Absolutely. If
- 45:06you're using AI. It's a thinking partner. Strictly
- 45:08as chatbot, you are missing the future. And that's
- 45:11my opinion, but. I don't know what a chatbot
- 45:13is. What is a chatbot? Chatbot, like chat GPT.
- 45:15I use chat GPT. Yeah, but that's. But what are
- 45:17you saying? It's more than that. That's why I'm
- 45:18saying. Like you actually, you can be a like
- 45:21investment partner. It can be a relationship
- 45:24partner. It can be. not relationship partner
- 45:26like you think not an actual partner but like
- 45:28it can help walk you through like she's being
- 45:30crazy like why is she being crazy and like well
- 45:32there's like based on her personality type right
- 45:34it's like 12 different things that she could
- 45:36be going through i have actually fed my therapist
- 45:38a relationship therapist i have actually programmed
- 45:42one of my gpts with all of my 10 years of research
- 45:45on mbti and the depth typology and all that other
- 45:48crap and it is like helped us actually work through
- 45:52relational like discourse and disconnects to
- 45:55understand how each person is viewing the perspective
- 45:57and it's been very interesting because we've
- 46:00actually been able to discover instances where
- 46:02we've essentially been like trigger spiraling
- 46:04very fascinating so that's been extremely helpful
- 46:07and that's something you wouldn't get more than
- 46:09likely out of a therapist because it's not tailored
- 46:12to specifically the way your mind works yeah
- 46:14so that's Honestly, the possibilities are limitless,
- 46:17in my opinion, with what you can do with AI.
- 46:19One of the big things I learned from it is it
- 46:21is tailored to how you think. Right. And if you
- 46:23can preempt it with profiles. You literally can
- 46:27actually have something that thinks like you
- 46:29think. I still need to build one for you. And
- 46:30looks for information how you look for information
- 46:32because I'm very analytical. I'm very data driven.
- 46:35The AI, once it gets to know me more and more
- 46:37and more, it's like it's going to. cut out all
- 46:39the emotional nonsense when it comes to investing.
- 46:43I just literally want the data so I can process
- 46:45the data and make decisions. And then if you're
- 46:47uncomfortable with making decisions, you can
- 46:50say, well, I used the data that you provided.
- 46:52I was thinking about investing in one of these
- 46:54three companies, which is the best company to
- 46:56invest in. You know how I think, you know the
- 46:58data I look at. The AI will actually pull that
- 47:00off and be like, well, this one's undervalued
- 47:01by this much. This one's about fair value. So
- 47:04your best bet of the three would be A. And that's
- 47:07the other thing too. If you have specific preferences
- 47:08like say value is the most important thing and
- 47:11then second is dividend growth and then third
- 47:12is whatever if you actually tell that that's
- 47:14your weighting system it'll literally do that
- 47:17synthesization for you so you don't have to actually
- 47:19like spend the time analyzing each individual
- 47:21one through that process you can have it do it
- 47:23and it'll shoot you out an actual rank system
- 47:25based on that hierarchy of your weighting metrics
- 47:28weighted metrics i guess i should say it's we
- 47:32use it a lot for a lot of different thought process
- 47:36thought and we put lots of variables into it
- 47:38the more variables the better your outcome right
- 47:41just like anything we use it as thought expansion
- 47:43so what it's basically doing is it's compressing
- 47:45the uh who said that the what the hell is his
- 47:49name peter lynch maybe the more stones you can
- 47:52turn over yeah so it's basically it's helping
- 47:54it's helping you turn over more more stones quicker
- 47:58because you just basically put in the detail
- 48:00like what stones you're looking for and it'll
- 48:01turn them over for you so like you it actually
- 48:03is really good to get started with ai even i
- 48:07mean i just used the generic google one i didn't
- 48:10even do like a right because we said we actually
- 48:12said back in the beginning that like if you don't
- 48:14even know which type of investing you're doing
- 48:16and you're just like looking for stocks you have
- 48:17no idea what you're even binding that by are
- 48:20you looking at s &p 500 funds are you using the
- 48:22bogle method do you want momentum trading are
- 48:24you doing the dividend approach are you doing
- 48:25like there's a whole assortment of different
- 48:27stuff and it's like If you can give it those
- 48:29information, it'll go in and look at those tickers
- 48:32and be like, that's irrelevant. So it rules it
- 48:33out for you. So you don't even have to take precious
- 48:35moments of your time to go and actually do that
- 48:38on your own. It's fantastic. Like, it's absolutely
- 48:40fantastic. You just have to learn to steer the
- 48:42tool. Don't be afraid of the tool. Steer the
- 48:44tool. Ideally, if I was new, I would play with
- 48:48it because we went over like the screening process
- 48:50that we look for. That screening process is going
- 48:52to take like the 60 ,000 things down to like
- 48:54a list of like 200, 300, 500, whatever. I would
- 48:57then literally take the 200, 300, 500, however
- 49:00many stocks are left over. I'd fucking copy a
- 49:02patient AI saying, here's the variables I'm looking
- 49:04for. Here's what a screener in Schwab or Fidelity
- 49:06gave me. So which ones are the best deals right
- 49:10now? And it'll process all of those. give you
- 49:13a much more comprehensive list that you would
- 49:15have maybe 20 or 30 i don't know that you would
- 49:19then have a you would have a plan going forward
- 49:21what to invest in because then you have to come
- 49:23up with your own plan i don't think i can help
- 49:24you with that like if you have ten thousand dollars
- 49:26and you wouldn't want to spread it over 25 stocks
- 49:28but you could in theory have a plan where you
- 49:31buy like four and then you just keep adding the
- 49:32ones on the list so long as they keep fitting
- 49:35your you have to come up with specific criteria
- 49:37for an allocation strategy which you need to
- 49:39steer it to give you the answer you really need
- 49:40which means you'd still need to do obviously
- 49:42some thought it's like to me i think it could
- 49:44be used as a secondary screener yes that'll do
- 49:46all the work for you Right. And it can actually
- 49:48help you get some hesitancy. It can help you
- 49:49do a lot of things more like quicker. And if
- 49:52you're hesitant because of taking like the human
- 49:54component out, one of my girlfriends was like,
- 49:56I do. She's like, I'm afraid to use AI. I finally
- 49:58got it out of her. She was like resistant, resistant,
- 50:00resistant. And I was like, OK, I was like, what's
- 50:02your issue here? Because you keep asking me the
- 50:03same question. You literally can like I built
- 50:06you this thing. Go use it. And she's like, well,
- 50:08I'm afraid of the like taking the human component
- 50:10out of it. And I was like, all right. I was like,
- 50:12that's a valid concern. But at the same time,
- 50:14if you never use it. You don't know if that's
- 50:16a valid fear, if that's just like some other
- 50:18misconstrued something or other. She started
- 50:20using it and she's like an artist. So it's been
- 50:22helping her get like really organized and ideas
- 50:24for things. And she's like, I understand why
- 50:26you love it now. And I was like, my point. I
- 50:28was like, so if you hate it, then don't use it.
- 50:30But if you don't use it and you don't know, just
- 50:33freaking try it. If you are afraid to start with
- 50:36the finance stuff, start it with something stupid.
- 50:39Like, help me reorganize my refrigerator. Help
- 50:42me plan my wardrobe. Just use it for something
- 50:45stupid that has no consequence and see what happens.
- 50:47You'll fall in love with it because, like, Jesus.
- 50:49That's literally a free assistant. Right. It's
- 50:51a free assistant. I've used it to help with tax
- 50:54returns. I've used it to help with freaking car
- 50:55repairs. I have used it. I've repaired HVACs,
- 50:58bike repairs. Bike training regimen. Health stuff.
- 51:03I mean, just insane things. You just have to
- 51:06know how to curb the dang thing. And trust me,
- 51:08I've hit walls with it. Got frustrated and I've
- 51:10had to build full AIs because I ran into an issue
- 51:13with the mechanic component. And it definitely
- 51:15led me astray. And then I was like, why did you
- 51:17tell me this bad information? And it told me.
- 51:19And then I basically broke. programmed one to
- 51:22avoid that problem. So like, if you have any
- 51:23questions about AI, just put it in the comments
- 51:25and she will actually answer it. Cause she's
- 51:27really good with AIs. Like she programs them
- 51:28and everything. I just use them. I don't do all
- 51:30the programming. She can actually help you foster
- 51:34a pretty good relationship with an AI that will
- 51:36be beneficial. But that's this week's podcast.
- 51:39I thought it was a really cool experiment with
- 51:41just AI do this, see what happens. And the AI
- 51:43just dumped out a bunch of shit. Then it dumped
- 51:45out a bunch of shit I didn't ask for. And then
- 51:46it dumped out other shit. I love when it gives
- 51:48you stuff you didn't ask for and you're like,
- 51:49well, that's cool. I asked why and it dumped
- 51:51out a bunch of shit. Like, well, here's the reason
- 51:52why. So they're basically programmed to give
- 51:55you as much information as possible. So the only
- 51:57limitation on information from AI is the limitation
- 52:00you put on yourself when asking the AI for stuff.
- 52:03Right. Much like life. Much like life. Much like
- 52:06life, but life's more complex and complicated,
- 52:09whereas AI is pretty simple. You just type in
- 52:11sentences. Yeah, that's the cool part. You can
- 52:14talk. I used to have issues. I have an engineering
- 52:16degree and I freaking sucked at computer programming.
- 52:18This is so easy for me. And even the guy that
- 52:21showed me this whole thing, he's like, wow, you
- 52:22picked up on this super quick because you're
- 52:23talking in like human language and it can interpret
- 52:25and like help you synthesize even your own speak
- 52:27to translate it into whatever the hell. Like
- 52:29I use it to build my own GPTs. Like it's, I know
- 52:33some would argue I'm not getting into like the
- 52:35full blown technical aspect of building, but
- 52:37they're useful tools for me. So I don't give
- 52:39a shit. Whatever works. All right, we will see
- 52:42you guys. See you next week. I think coming up,
- 52:44we're going to have our May cash flow coming
- 52:46up next. Yeah, I got to do that soon. But we
- 52:48also probably do the next weekly before. So I
- 52:50still got to edit. Oh, yeah, that weekly is coming
- 52:52up. We're back on the road heading up to South
- 52:54Dakota. So it's going to be a little and we're
- 52:56trying to survive a heat wave. So we actually
- 52:57just decided on a whim. We're staying here for
- 52:59like three extra days. Yeah, the weather changed
- 53:02like last week. It was supposed to be decent
- 53:04like 80s during the day. And it went from the
- 53:0680s to like three days over 100 degrees. So we're
- 53:08just going to be hanging out in this forest.
- 53:09And we ran out of cat foods. Eating them like
- 53:11these chicken packet things we have and tuna.
- 53:13So we're going to be hiking and biking around
- 53:14this area. There's worse places to be. I haven't
- 53:17seen a person for over 24 hours. That's so nice.
- 53:21So we'll see you guys, I guess, later in the
- 53:23week when we do the weekly. Right -o.