00:00:04
Okay, today I want to paint a sort of
00:00:06
picture for everyone about the current
00:00:08
state of AI technology. Not because I
00:00:10
don't think it's dangerous. It most
00:00:11
certainly is. Or that it won't have an
00:00:13
outsized impact on various different
00:00:14
career fields because it will with
00:00:16
likely insane levels of consequence, by
00:00:18
the way, but more so because the current
00:00:20
state of the industry when it comes to
00:00:21
startup companies is kind of hilarious.
00:00:24
Let me begin with this sort of
00:00:26
foundation narrative building, you could
00:00:28
call it. In March of 2019, which doesn't
00:00:30
seem that long ago on paper, but in the
00:00:32
world of technology is rather
00:00:33
substantial, British venture capital
00:00:35
firm MMC conducted a survey of 2830
00:00:39
European startups that are officially
00:00:40
classified as AI companies. What's
00:00:43
interesting about that, however, is that
00:00:45
only 1580, 1,580 of those companies were
00:00:48
found to have any evidence of actually
00:00:50
working in the field, while the
00:00:52
remaining 1,250, which is about 40% of
00:00:55
the entire group, probably had nothing
00:00:57
to do with it. This was a
00:00:59
two-dimensional problem at the time.
00:01:01
First, because some of the startups were
00:01:02
obviously pretending to have AI
00:01:04
relevance, capitalizing on hype, but
00:01:06
second, and more importantly, because
00:01:08
even financial analysts were
00:01:10
mclassifying companies as AI related,
00:01:13
even when they were not in fact related
00:01:15
in the slightest. Quote, possibly
00:01:17
because AI companies attract higher
00:01:19
investments. End quote. See the problem?
00:01:22
Obviously, this particular research is
00:01:23
based in Europe, but 40% of startup
00:01:26
companies being mclassified as AI based
00:01:28
when they actually have nothing to do
00:01:30
with it, likely for the purpose of
00:01:31
attracting higher investment. That's an
00:01:33
issue. And it's probably not all that
00:01:35
different here in America either. Except
00:01:37
why don't we go one step further?
00:01:39
Instead of focusing on companies that
00:01:41
obviously have nothing to do with AI,
00:01:43
yet still classify themselves or get
00:01:44
classified by analysts as being
00:01:46
AIcentric, why don't we turn and focus
00:01:48
on a very much not uncommon situation
00:01:51
where companies that seem to be well
00:01:52
positioned in the world of AI
00:01:54
development or adoption are simply lying
00:01:57
about it. Of course, everyone saw the
00:01:59
title. So, let's dig into the reality of
00:02:01
AI being actually Indian workers. The
00:02:05
biggest and most recent example of this,
00:02:07
but trust me, there's many more, is a
00:02:09
company called Builder.ai.
00:02:11
Supposedly leveraging a combination of
00:02:12
human workers and AI language models
00:02:14
with extreme emphasis placed on the
00:02:16
technology side of things. Builder.ai
00:02:18
positioned itself as a $ 1.5 billion
00:02:21
Microsoftbacked startup. They had this
00:02:24
virtual assistant called Natasha, which
00:02:26
would supposedly help people build
00:02:28
software applications. But in June of
00:02:30
2025, after a creditor seized something
00:02:32
like $37 million from their bank account
00:02:35
as a result of poor financial
00:02:36
performance, the world suddenly found
00:02:38
out that no, builder.ai was not some
00:02:41
clever use of technology. It was instead
00:02:44
about 700 to a,000 workers from India
00:02:47
pretending to be chatbots. Digging a
00:02:50
little bit deeper, the writing was on
00:02:51
the wall for this company for about 5 or
00:02:53
6 years beforehand. All the way back in
00:02:56
2019 when the company was known as
00:02:58
Engineer.ai. AI. The Wall Street Journal
00:03:00
reported, quote, "Startupngineer.ai
00:03:03
says it uses artificial intelligence
00:03:05
technology to largely automate the
00:03:06
development of mobile apps, but several
00:03:08
current and former employees say the
00:03:10
company exaggerates its AI capabilities
00:03:12
to attract customers and investors." End
00:03:15
quote. That's basically a six-year
00:03:17
premonition of what would ultimately
00:03:18
destroy the company, seeing as they've
00:03:20
now entered bankruptcy proceedings. But
00:03:22
when a $ 1.5 billion startup backed by
00:03:25
Microsoft goes belly up for faking the
00:03:27
entire premise of their AI technology
00:03:30
instead using human workers, it requires
00:03:33
the question to be asked, how many more
00:03:35
companies are doing this? Answer, a lot
00:03:38
of them are doing it. And if we think of
00:03:39
the entire industry right now, not only
00:03:41
are nearly 40% of AI startups being
00:03:44
inaccurately classified in the first
00:03:46
place for the purpose of attracting
00:03:47
undeserved investment, the remaining
00:03:49
companies that aren't might very well be
00:03:52
faking it. Let's use another example.
00:03:54
Pretty much all of the world's largest
00:03:56
companies are actively developing their
00:03:58
own AI models. Think Apple, Google,
00:04:01
Microsoft, etc. But one company in
00:04:03
particular, Amazon in this case, is
00:04:05
where I want to focus. Amazon is
00:04:07
primarily an online e-commerce cloud
00:04:10
computing and shipping company, but they
00:04:12
also have a very specific small retail
00:04:14
offering called Just Walk Out
00:04:17
Technology. Reading from the company's
00:04:18
own website, quote, "Just walk out
00:04:21
eliminates traditional checkout lines,
00:04:23
creating a seamless shopping experience
00:04:25
using advanced AI, sensors, computer
00:04:28
vision, and RFID. Just Walk Out
00:04:30
accurately tracks item selection and
00:04:32
automates payment when shoppers exit the
00:04:34
store." End quote. That's pretty cool on
00:04:37
its face, right? Walk in, do your
00:04:39
shopping, walk out, and the AI takes
00:04:41
care of everything else. It's a
00:04:43
nightmare in terms of privacy, but it's
00:04:44
a benefit in terms of convenience, which
00:04:46
most people are happy to accept as a
00:04:47
trade. However, come to find out, first
00:04:50
reported by the information, Just Walk
00:04:53
Out was largely reliant, as in about 70%
00:04:56
of all orders, on human workers in India
00:05:00
who would physically watch what
00:05:01
customers picked up or walked out with
00:05:04
on some sort of feed presumably and
00:05:06
verify the orders that way. Once again,
00:05:09
revolutionary AI technology was just a
00:05:12
bunch of lowpaid overseas workers. A
00:05:15
couple years later, Amazon began moving
00:05:17
towards something called Dash Carts,
00:05:19
which would detect whatever item you
00:05:20
physically put in your cart and make
00:05:22
checkouts easier that way. But the point
00:05:24
I'm trying to make is that Dash carts
00:05:26
don't have advanced AI technology and
00:05:28
computer vision. It's just a scanner
00:05:31
system that replaces the checkout
00:05:32
counter. That's two examples so far, but
00:05:35
next one is by far my favorite. So, let
00:05:38
me just play a news clip for you real
00:05:40
fast. Hi, what can I make fresh for you
00:05:42
today? AI is here in the valley and it's
00:05:44
looking to take your order. What else?
00:05:46
It's a new system. It's reliable. It's
00:05:48
there 24/7 and it really helps. In
00:05:51
partnership with Presto Automation,
00:05:53
Carl's Jr. is bringing out artificial
00:05:55
intelligence to take orders at
00:05:56
drive-throughs. With one of the first
00:05:58
locations in the country being in
00:06:00
Buckeye, okay, so AI partnership for
00:06:03
takeout orders, Presto the company, and
00:06:05
Carl's Jr. as the customer. Sounds good.
00:06:08
Well, Presto is an AI classified startup
00:06:11
company providing their automated
00:06:13
solutions to fast food giants and chains
00:06:16
with customers like Del Taco, Hardies,
00:06:19
Checkers, and obviously Carl's Jr.,
00:06:20
right? That one we already said. But if
00:06:22
you start looking at the way that their
00:06:24
executives talk about the service,
00:06:26
something really strange happens. In
00:06:28
late 2021, December to be precise,
00:06:31
speaking to Karen Webster of a site
00:06:33
called Payments, CEO of Presto, Rajat
00:06:36
Rajuri had this to say. quote, "Our
00:06:39
system actually takes over 95% of all
00:06:43
orders without human intervention. We're
00:06:45
about to announce the largest national
00:06:47
rollout of this technology ever. It's
00:06:50
going to be announced in early January.
00:06:51
It's going to be with a big chain and
00:06:53
they're going allin on it." End quote.
00:06:56
Sounds great, right? AI tech startup,
00:06:58
the biggest rollout ever with a major
00:07:00
partner. 95% of orders require no human
00:07:04
intervention, which are very carefully
00:07:07
chosen words, by the way. But we have to
00:07:08
keep in mind that there is a very big
00:07:11
difference in what the executives of a
00:07:12
company like this are willing to say to
00:07:14
the media, compared to what they're
00:07:16
willing to say when it's the SEC. Yeah.
00:07:19
If you switch over and look at a form 8K
00:07:21
Presto Automations November 16th, 2023
00:07:24
filing with the SEC, we find a very
00:07:27
interesting statement. Quote, "As we
00:07:30
continue to grow our Presto Voice
00:07:32
business, we are focusing on increasing
00:07:34
the rate at which our solution
00:07:35
successfully receives orders without the
00:07:37
need for any human agent intervention,
00:07:40
commonly referred to as humans in the
00:07:42
loop. Currently, over 70% of orders
00:07:45
taken by our Presto Voice solution
00:07:47
require human agent intervention. As we
00:07:50
continue to improve our AI accuracy and
00:07:52
further deploy Presto Voice across store
00:07:54
locations, we believe that the
00:07:56
percentage of orders that do not require
00:07:58
any human agent intervention will reach
00:08:00
30% or better. End quote. Let's get
00:08:03
something straight here. Okay. In 2021,
00:08:07
Mr. Suri was telling the media that 95%
00:08:10
of their orders are entirely AI
00:08:12
processed. However, in 2023, his company
00:08:15
was telling the SEC that less than 30%
00:08:18
of their orders were automated, over 70%
00:08:22
required direct human intervention to
00:08:24
make them accurate, and that the company
00:08:25
was hopeful that they might achieve a
00:08:28
less than one-third automation rate in
00:08:31
the future. I'm not going to make any
00:08:32
sort of definitive statements here, but
00:08:34
that kind of feels like it might be
00:08:36
outright fraud. Here's the hook.
00:08:38
Speaking of fraud, we have another great
00:08:40
example. This time from an AI e-commerce
00:08:43
checkout company called Nate.te.
00:08:46
Nate.te is still online, though it
00:08:49
probably shouldn't be. This is their
00:08:50
website right here advertising a
00:08:52
supposed mobile application where all
00:08:54
you have to do is click something and
00:08:56
the entire order gets processed for you
00:08:58
with an AI solution. It's a timesaver.
00:09:01
Once again, convenience is the selling
00:09:02
point by avoiding the tedium of entering
00:09:04
payment credentials, billing details,
00:09:06
and personal information on whatever
00:09:08
various platforms you shop under, which
00:09:10
is very sensitive information, I might
00:09:12
add. However, in April of 2025, handed
00:09:16
down in the Southern District of New
00:09:18
York, a fraud indictment became public,
00:09:20
alleging, quote, "Albert Sager defrauded
00:09:23
investors with fabrications of his
00:09:25
company's purported artificial
00:09:26
intelligence capabilities while covertly
00:09:28
employing personnel to satisfy the
00:09:30
illusion of technological automation."
00:09:33
End quote. That's already pretty
00:09:34
succinct. And we have to remember here
00:09:36
that there's a massive implication
00:09:38
behind the process of giving payment
00:09:39
details to a company and then having it
00:09:41
not actually be an automated solution
00:09:43
that processes your payment details.
00:09:45
Okay, there's a whole quagmire of
00:09:47
problems there. But let me just read
00:09:49
another paragraph for editorial
00:09:50
emphasis. I'll call it quote. Sager used
00:09:54
hundreds of contractors or purchasing
00:09:56
assistants in a call center located in
00:09:58
the Philippines to manually complete
00:10:00
purchases occurring over the Nate app.
00:10:02
In or about the fall of 2021 with the
00:10:05
busy holiday shopping approaching and
00:10:07
despite his numerous prior
00:10:08
representations that Nate did not use
00:10:10
bots or dumb bots as he referred to
00:10:12
them, Sager directed Nate's engineering
00:10:14
team to develop bots to automate some
00:10:17
transactions on the Nate app. After
00:10:19
creating the bots, Nate used bots in
00:10:21
addition to the manual teams to complete
00:10:23
purchases that were purportedly being
00:10:26
completed by AI technology. End quote.
00:10:29
That is absolutely hilarious to me
00:10:31
because it perfectly exemplifies a much
00:10:33
larger thanex expected portion of the AI
00:10:36
tech bro startup world. This CEO
00:10:39
decided, let's make an AI application
00:10:42
and then failed to do so. instead using
00:10:44
underpaid human workers in the
00:10:46
Philippines. After which he returned to
00:10:48
prior existing technology which he had
00:10:50
previously referred to as dumb using a
00:10:52
combination of those dumb bots and low
00:10:55
paid human workers all for the purpose
00:10:57
of impersonating an AI program. Oh, and
00:11:01
the company attracted $51 million across
00:11:04
three separate funding rounds. Let's get
00:11:06
that straight. How about a recap with
00:11:09
just these four examples so far?
00:11:11
Billions of dollars in investment
00:11:13
capital has been funneled into companies
00:11:15
that would supposedly use AI to
00:11:17
revolutionize app development, retail
00:11:19
checkouts, e-commerce checkouts, and
00:11:21
fast food ordering, all of which are
00:11:23
substantial businesses. Builder.ai for
00:11:26
the app development side was just a
00:11:28
bunch of underpaid developers in India.
00:11:31
Amazon's just walk out technology,
00:11:33
retail checkouts this time, now seems to
00:11:35
be getting pushed out the door, no pun
00:11:37
intended there, after allegations that
00:11:39
instead of advanced machine learning, it
00:11:41
was just overseas workers also in India.
00:11:44
Presto, the fast food delivery
00:11:46
disruptor, if you want to call it that,
00:11:47
is on record stating that 95% of orders
00:11:50
have been successfully automated since
00:11:52
2021, but then quietly to the SEC
00:11:54
clarifying that not even onethird of
00:11:57
their orders have been automated in
00:11:58
2023, with all the rest requiring human
00:12:01
intervention. and Nate.te supposedly
00:12:04
revolutionizing the e-commerce checkout
00:12:06
process with artificial intelligence and
00:12:08
major financial backers was again really
00:12:11
just a bunch of human workers this time
00:12:13
in the Philippines. Pretty
00:12:15
impressive industry we got going on here
00:12:16
guys. I want to end with this. I am
00:12:19
fully aware of the dangers that AI
00:12:20
automation can pose. I think there will
00:12:22
be incredibly harsh consequences for a
00:12:24
lot of companies, career fields, and
00:12:26
white collar workers especially. And I'm
00:12:28
not sure what regulatory responses would
00:12:30
be appropriate right now or if an
00:12:32
adequate response will even ever appear.
00:12:34
But I do also understand that the
00:12:36
industry is now unbelievably bloated.
00:12:39
Artificial intelligence across multiple
00:12:41
sectors is a lightning rod for
00:12:43
multi-billion dollar fraud schemes and
00:12:45
self-enrichment of the highest order.
00:12:47
The level of delusion being pedled
00:12:49
widespread now and subsequently eaten up
00:12:51
by venture capital investment firms
00:12:53
among many others is staggering. And
00:12:56
even if the underlying technology does
00:12:57
indicate massive upheaval in the future,
00:13:00
the current state of the AI industry is
00:13:02
a clear and obvious bubble. What we see
00:13:05
today is unsustainable in any sense of
00:13:08
the word when it comes to cash infusion
00:13:10
with a super majority of all companies
00:13:12
operating in the space destined for
00:13:14
total failure because nearly half of
00:13:16
them first of all don't as in don't
00:13:18
operate in the space to begin with that
00:13:21
is. and the remaining portion of the
00:13:23
remaining portion rather it's infested
00:13:25
with outright fraud. In the end, AI is
00:13:28
certainly a revolutionary technological
00:13:30
concept when perfected or massively
00:13:32
progressed. But the AI industry as it
00:13:34
stands is just the next iteration of
00:13:36
crypto, the dot bubble, or even Dutch
00:13:39
tulip mania. 40% joke, 30% fraud, 20% I
00:13:44
don't even know what's happening right
00:13:45
now, and maybe 10% real at best. Maybe.
00:13:50
That's it. If you want to support the
00:13:51
channel, check out the links down below.
00:13:52
A special VPN deal for anyone
00:13:54
interested. Great one there. Locals and
00:13:56
Patreon for monthly memberships,
00:13:57
merchandise, social media, etc., etc.
00:14:00
But I'll cut it there and stop rambling.
00:14:01
As always, thank you all for watching.
00:14:03
Question everything and have a nice
00:14:05
night.
