Has Google Met its Match | ChatGPT and Perplexity AI

Best of Us Investors
6 Mar 202413:40

Summary

TLDRThe video script discusses the rise of conversational AI search engines like Perplexity AI, challenging Google's dominance. It compares Perplexity AI with ChatGPT, highlighting Perplexity's ability to provide direct answers through natural language interaction, access real-time data, and engage in contextual conversations. The script suggests that tech giants like Amazon, NVIDIA, and investors like Jeff Bezos are taking notice and investing in this space, as the traditional search model is becoming outdated. It encourages viewers to ask insightful questions and stay curious about emerging startups and technologies that could disrupt the search engine landscape.

Takeaways

  • 😃 The traditional way of searching on Google provides many web pages as results but not a direct answer, while AI models like ChatGPT and Perplexity AI aim to provide more conversational and direct answers.
  • 🤖 ChatGPT excels at tasks like data analysis, image generation, and answering follow-up questions based on previous responses, but its knowledge is limited to a certain cutoff date.
  • 🔍 Perplexity AI is a newer AI model that can access and analyze the latest data in real-time, providing more up-to-date and contextual answers through a conversational interface.
  • 💰 Major tech companies like Amazon, Nvidia, and Microsoft are investing in AI startups like Perplexity AI, recognizing the potential shift in search and AI technologies.
  • 🎯 To get the most out of these AI models, users need to improve their ability to ask specific, in-depth questions and provide proper context for the desired information.
  • 🌐 The rise of AI models like Perplexity AI could potentially disrupt Google's dominance in the search market if they fail to effectively integrate these new technologies.
  • 📈 Investors should keep an eye on small AI startups and the companies investing in them, as they may become acquisition targets or major players in the evolving search and AI landscape.
  • 💡 The script encourages curiosity, asking questions, and exploring new technologies to uncover potential opportunities and innovations.
  • 📚 The script promotes the author's paid subscription service, offering access to additional resources and insights related to investing and emerging technologies.
  • 🔑 Effective questioning and contextual prompting are key to maximizing the capabilities of AI models like ChatGPT and Perplexity AI.

Q & A

  • What is the main topic of the video script?

    -The main topic of the video script is the emergence of new AI-powered search engines like Perplexity AI, and how they challenge the dominance of traditional search engines like Google by providing more conversational and direct answers.

  • What are the key differences between ChatGPT 4 and Perplexity AI?

    -ChatGPT 4 is not a full-fledged search engine like Google and Perplexity AI. It excels at analyzing data and creating images, but it operates on a fixed knowledge base that may not be entirely up-to-date. Perplexity AI, on the other hand, is a conversational search engine that can access and provide answers based on the latest data in real-time.

  • How does Perplexity AI work in terms of providing answers?

    -Perplexity AI works more like a human conversation, where it provides direct answers to questions and prompts users to ask follow-up questions for further clarification or elaboration. It doesn't simply provide a list of potential sources like traditional search engines.

  • What is the significance of the investment from companies like Amazon, Nvidia, and Jeff Bezos in Perplexity AI?

    -The investments from major tech companies and influential figures like Jeff Bezos indicate that the smart money recognizes the potential of conversational AI and new search technologies like Perplexity AI to disrupt the traditional search engine landscape dominated by Google.

  • What is the key to getting the most out of AI search engines like Perplexity AI and ChatGPT 4?

    -The key to getting the most out of these AI search engines is the ability to ask in-depth, specific questions and prompts. By providing more context and details in our queries, we can guide the AI to provide more tailored and insightful answers.

  • What is the potential impact of companies like Perplexity AI on Google's dominance in the search engine market?

    -The script suggests that Google's dominance in the search engine market may be threatened by the rise of conversational AI and companies like Perplexity AI that can provide more direct and up-to-date answers. It compares Google's potential situation to companies like Kodak, which failed to adapt to new technologies and lost market share.

  • What is the advice given in the script regarding investments in this emerging field?

    -The script advises keeping an eye on small startup companies and companies investing in them, as they may be the ones developing the next level of generative AI and search technologies. It suggests that these companies could be potential acquisition targets for larger players like Amazon, Nvidia, and others.

  • How does the script describe the process of creating thumbnails using ChatGPT 4?

    -The script mentions that the thumbnails used in the videos are created by ChatGPT 4 through a series of prompt questions. The AI generates images based on the prompts, which may require multiple iterations to get the desired look.

  • What is the significance of the comparison between Perplexity AI's team size (36 people) and its potential impact?

    -The script highlights that Perplexity AI, with just 36 people, has figured out how to use generative AI and reinforcement learning to create a conversational search engine that can challenge Google's dominance. This small team size is contrasted with the potential disruption they could bring to the search engine market.

  • What is the key takeaway or call to action from the script?

    -The key takeaway and call to action from the script is to be curious, ask better questions, and be on the lookout for small companies and startups working on emerging technologies like conversational AI and generative search. These companies could represent potential investment opportunities and disruptions to established players like Google.

Outlines

00:00

🔍 The Search for Better Search

The paragraph discusses the evolution of search engines and the limitations of traditional search methods like Google, which provide numerous links but no direct answers. It introduces perplexity AI, a new player in the search field that aims to provide natural language responses through generative AI and reinforcement learning. The paragraph also mentions the involvement of companies like Microsoft, SK Telecom, Amazon, and Nvidia in the development of perplexity AI.

05:01

🤖 The Differences Between AI Assistants

This paragraph compares the capabilities of ChatGPT 4 and perplexity AI. It highlights that ChatGPT 4 excels in data analysis, image creation, and can generate content based on prompts, while perplexity AI is designed as a conversational search engine that provides direct answers through natural language interactions. The paragraph also mentions perplexity AI's ability to access the latest data, unlike ChatGPT 4, which has a cutoff date for its knowledge base.

10:01

🌟 The Future of Search and Investing

The paragraph explores the potential implications of generative AI and companies like perplexity AI on the future of search. It suggests that traditional search methods like Google may become obsolete, and companies that embrace conversational AI and natural language search could disrupt the market. The paragraph encourages investors to keep an eye on small startups and companies investing in generative AI, as they may be acquired by larger players. It also emphasizes the importance of asking insightful questions and prompts to get the most out of these AI assistants.

Mindmap

Keywords

💡Generative AI

Generative AI refers to artificial intelligence systems that can generate new data, such as text, images, or audio, based on the data they were trained on. In the video, generative AI models like ChatGPT and Perplexity AI are discussed as alternatives to traditional search engines, as they can provide direct answers to queries instead of just links. The video mentions how ChatGPT can generate images based on prompts, and how Perplexity AI aims to have more natural language conversations.

💡Search Engine

A search engine is a software system designed to search for information on the internet or within a specific database. The video contrasts traditional search engines like Google, which provide links and resources to find answers, with newer models like Perplexity AI that aim to directly provide answers through natural language conversations. The ability of Perplexity AI to access and process real-time data is highlighted as an advantage over models like ChatGPT, which have a fixed knowledge cutoff date.

💡Natural Language Processing (NLP)

Natural Language Processing (NLP) is a branch of artificial intelligence that deals with the interaction between computers and humans using natural language. The video discusses how models like Perplexity AI and ChatGPT use NLP to understand and respond to queries in a more human-like manner, engaging in conversational exchanges instead of just providing links or resources.

💡Reinforcement Learning (RL)

Reinforcement Learning (RL) is a type of machine learning technique where an AI system learns by interacting with an environment and receiving rewards or punishments for its actions. The video mentions that Perplexity AI utilizes generative RL, which likely refers to combining generative models with reinforcement learning techniques to improve the system's ability to generate relevant and coherent responses.

💡Monopoly

A monopoly refers to a situation where a single company dominates a particular market or industry. The video suggests that Perplexity AI and other emerging search technologies could potentially disrupt Google's long-standing monopoly in the search engine market by offering a more advanced and conversational search experience.

💡Data Analysis

Data analysis refers to the process of inspecting, cleaning, transforming, and modeling data to discover useful information, draw conclusions, and support decision-making. The video highlights ChatGPT's ability to analyze data from spreadsheets and help businesses gain insights, suggesting that advanced AI models can assist with complex data analysis tasks.

💡Prompting

Prompting refers to the act of providing an initial input or instruction to a language model or AI system to guide its response or generation. The video emphasizes the importance of skillful prompting when interacting with models like ChatGPT to obtain desired outputs, such as generating specific images or answering complex queries.

💡Kodak Moment

The term Kodak moment is used in the video as a metaphor for a company that fails to adapt to changing technology or market conditions, leading to its downfall. The video questions whether Google's struggle to integrate generative AI effectively into its search engine could be a Kodak moment, where it loses market share to more innovative competitors.

💡Startups

Startups are newly established companies, often focused on developing innovative products or services. The video suggests keeping an eye on small startups working on generative AI and search technologies, as they could potentially be acquired by larger companies like Amazon, NVIDIA, or others interested in this space.

💡Venture Capital

Venture capital refers to the investment of money into startup companies or small businesses with the potential for high growth. The video mentions companies like Microsoft, Amazon, and NVIDIA investing in startups like Perplexity AI and Anthropic (creators of ChatGPT), highlighting the interest and capital being poured into the development of generative AI and search technologies.

Highlights

Until recently, when asking a question on Google, we would get thousands of pages of answers or links, but not the direct answer.

ChatGPT and other AI models like GPT-4 can now provide direct answers to questions, analyze data, and create images, changing the way we search for information.

A new player in the search world is Perplexity AI, which uses natural language processing to have a conversation-like interaction and provide direct answers.

Perplexity AI has signed a deal with SK Telecom, potentially becoming an alternative to Google for customer search.

Perplexity AI can access the latest data in real-time, while ChatGPT has a cutoff date for its knowledge base.

Companies like Amazon, Nvidia, and Jeff Bezos are investing in Perplexity AI, recognizing the potential disruption in the traditional search model.

The key to effectively using AI search assistants like Perplexity AI and ChatGPT lies in our ability to ask in-depth, specific questions.

Prompting ChatGPT effectively, such as by providing context and constraints, can lead to more accurate and tailored results.

Google may be facing challenges in incorporating generative AI into its search engine, potentially losing market share to innovative startups.

Investors should keep an eye on small startups and companies investing in them, as they could be acquired by larger players in the AI search space.

Becoming better at asking questions and being curious is crucial for finding valuable insights and opportunities in the evolving AI search landscape.

ChatGPT's image creation capabilities allow for generating custom thumbnails through iterative prompting.

Data analysis with ChatGPT, such as uploading spreadsheets and asking questions, can lead to business improvements and cost savings.

Perplexity AI's conversational approach to search mimics how humans interact, potentially providing a more natural and intuitive experience.

The future of search may lie in small, innovative companies leveraging generative AI and reinforcement learning to disrupt the traditional model.

Transcripts

play00:02

when you ask your spouse the question

play00:05

what Bowls do we use to make uh bread

play00:08

pudding in what how do they answer that

play00:11

question do they come back with oh we

play00:14

use those uh you know plastic bowls down

play00:18

there on the right hand side just use

play00:21

the bigger one because it you know makes

play00:23

a mess or do they come back

play00:26

with a bunch of different references

play00:29

that you could go

play00:30

to and try to figure out where that bowl

play00:32

is and what type of Bowl it is well

play00:35

that's what we have grown used to when

play00:37

it comes to search when I have a

play00:41

question and I want to find out an

play00:44

answer I go to Google right and we have

play00:47

all started going to Google for many

play00:50

many years and Google is the number one

play00:53

most dominant used search engine on the

play00:57

planet up until recently

play01:00

when I go to Google and I ask a question

play01:04

I get thousands of pages of answers that

play01:08

could you know links that could answer

play01:10

my question what I don't get is an

play01:14

answer well that all sort of started

play01:16

changing about a year and what five six

play01:20

months ago with chat GPT 4 or chat

play01:24

gpt3 then four and then so on and so

play01:27

forth there is a new play here in the

play01:30

search world and that is perplexity

play01:36

AI I came across perplexity AI after

play01:41

really doing some research on the whole

play01:44

debacle that J Google had with uh Gemini

play01:48

in their image creating uh AI program

play01:52

and since has been taken off line and

play01:55

being fixed because of its bugs well you

play01:59

if you understand the agenda that a very

play02:02

big part of Google has had when it comes

play02:06

to diversity and all this it's sort of

play02:08

clouding their vision of what their

play02:11

business is all about well companies

play02:14

such as chap GPT or open AI who got an

play02:18

enormous amount of money from

play02:22

Microsoft about a year or so ago or yeah

play02:26

about a year or so ago uh I think it was

play02:28

10 billion if I remember correctly it's

play02:30

a lot of money it has really started to

play02:33

open up the field of generative Ai and

play02:37

that is artificial intelligence that

play02:41

learns that starts to speak in un and

play02:44

I's natural language well there is a new

play02:47

player in town that's perplexity Ai and

play02:51

recently I se article on Reuters with

play02:54

them about how they just signed a deal

play02:57

with SK Telecom about being a option to

play03:02

Google when it comes to uh their

play03:05

customer search which got me to thinking

play03:08

what is this perplex AI I mean this

play03:12

could be pretty interesting this could

play03:14

actually basically break up the Monopoly

play03:18

that Google has on search and see the

play03:21

the problem Google has had is

play03:24

incorporating artificial intelligence or

play03:26

generative artificial intelligence into

play03:29

their program now you can go to Google

play03:31

today and you can type in a question and

play03:35

it will give you its traditional list of

play03:38

resources to go figure out the answer

play03:41

for yourself but it also has a little

play03:44

section at the

play03:45

top where it's their AI version or

play03:49

Gemini and it's you know it's all text

play03:51

it's no image creation so I got to

play03:54

thinking about all this and I thought

play03:56

hey what is the differences

play04:00

between Google

play04:02

cpt4 in particular and perplexity Ai and

play04:07

so this is what I asked proxity Ai and

play04:11

basically it came back with a handful of

play04:13

answers first of all jet chat

play04:17

gp4 is not necessarily a full-on search

play04:21

engine like Google is and perplexity so

play04:26

what chat GPT 4 does very well is it can

play04:30

analyze data you can upload a

play04:33

spreadsheet and as start asking it

play04:35

questions about that spreadsheet and

play04:38

analyze that data which I have a friend

play04:40

who has done that and is now changing

play04:43

his whole business and saving him a

play04:46

whole lot of money when it comes to data

play04:48

analysis but also can create images if

play04:52

you watch any of my videos you've seen

play04:55

the thumbnails that I've been using well

play04:57

that's chat gp4 for creating those

play05:01

thumbnails for me through a number of

play05:03

prompt questions that I continuously ask

play05:06

to get the look I want now it's not

play05:09

perfect it can't spell worth a crap

play05:12

which makes me feel comfortable because

play05:14

I can't spell either so it's nice to

play05:17

know that it hadn't figured out how to

play05:19

spell either so it's great but what it's

play05:22

interesting is how I go about asking

play05:26

Chachi PT for questions how I prompt it

play05:31

to produce the answers that I'm looking

play05:35

for well perplexity does very similar

play05:39

thing what's unique about this is that

play05:43

perplexity is one of those natural

play05:47

language uh programs where like you and

play05:50

I are talking you ask me a question and

play05:53

I give you an answer whereas if I ask

play05:57

Google a question it gives me

play06:00

thousands of pages of potential answers

play06:03

but never the answer whereas perplexity

play06:07

gives me an answer and then it prompts

play06:09

me to IND you know ask more about that

play06:15

question I just asked and the answers

play06:17

that it just gave so it's like a

play06:20

conversation that you and I are having

play06:23

about a subject what it doesn't do it

play06:26

doesn't create

play06:28

images um from what I can tell it

play06:31

doesn't analyze data it is truly a

play06:35

search engine that works more like a

play06:38

human being does and how you and I would

play06:41

interact so what's interesting about

play06:44

this is is that prxi is at the time of

play06:48

the research I've found it's like 36

play06:51

people it's 36 people who have figured

play06:55

out how to use generative Ai and RL uh

play07:01

uh AI or R generative RL which is

play07:04

reinforced

play07:05

learning and what it basically does is

play07:08

it's they're creating this platform for

play07:12

which I can go ask questions as I would

play07:15

be asking you and it can go ask access

play07:19

all the data in the world in real time

play07:22

that's one of the drawbacks of cat GPT 4

play07:25

and three is that it's data that's up to

play07:29

a certain period of time so for instance

play07:32

before I believe it's up to March of uh

play07:38

203 so it's not the latest data whereas

play07:42

perple perplexity is latest data it's up

play07:46

to speed now it's not a publicly traded

play07:49

company yet um in honestly I think

play07:53

somebody's going to come in and buy this

play07:55

thing up who well who could be

play08:00

Amazon because Amazon's investing money

play08:03

into them it could be Nvidia because

play08:07

they've invested money into them and of

play08:09

course one of the biggest investors out

play08:13

there right now who's putting money in

play08:15

all different things like um uh what's

play08:18

the robot company figure AI um and he

play08:24

just recently sold off 50 million shares

play08:27

of Amazon Jeff Bezos is investing into

play08:30

this company as well so what this tells

play08:33

me is the Smart money

play08:35

recognizes that your traditional way we

play08:38

have gone about doing

play08:41

search is Now quickly becoming a

play08:44

conversation between me and my 10

play08:47

zillion friends who are really smart and

play08:52

being able to decipher all that

play08:54

information and bring it down to one

play08:56

answer and then to be able to elaborate

play08:59

on that answer as I prompted to ask

play09:02

better or more questions about that data

play09:06

now here's where things are going to get

play09:10

differentiate their selves and that's on

play09:12

you and

play09:13

I and that is our ability to ask indepth

play09:19

deep questions that can give

play09:22

us the answers that we're looking for

play09:26

now if I just said you know how do I

play09:28

make a lemon

play09:30

pie it they all will give me an answer

play09:33

now Google will give me 10,000 pages of

play09:36

answers but they all give me an answer

play09:38

but if I ask how do I make a lemon pie

play09:42

in my apartment and I only have a um

play09:48

stove top no oven how would I go about

play09:51

doing that and I have limited resources

play09:55

to get ingredients what's the best way

play09:57

to do it first of all I don't I don't

play09:59

know if I'd ask if I wanted a lemon pie

play10:01

or how to cook a lemon pie I that would

play10:03

sounds terrible but it could give me a

play10:06

more in-depth answer of how I would do

play10:10

that so it's on me to start getting

play10:12

really good at asking questions or

play10:15

prompts so for instance with cha

play10:20

gp4 I'll go in and say hey you are a

play10:24

graphic artist who just got assigned a

play10:28

uh project for a

play10:30

financial uh educational website that

play10:34

talks about stocks and markets and

play10:37

Global

play10:38

economics and we just he just did a

play10:41

video on the differences between Google

play10:45

chap G pd4 and perplexity

play10:49

AI give me a thumbnail that has these

play10:53

colors in it and give this imagery and

play10:57

guess what it goes out and creates it

play10:59

now does it get it right the first time

play11:01

no it doesn't but I would have never

play11:03

gotten that on a Google Search and this

play11:07

is where I start to

play11:09

wonder is

play11:11

Google a Kodak is

play11:16

Google uh a company who has reigned

play11:21

Supreme for 2ome years but is now

play11:26

fumbling because of management MISD

play11:29

ction or misguidance or lack of guidance

play11:33

and is Google starting to go the other

play11:37

side of the bell curve and will they

play11:40

lose market share to companies such as

play11:44

perplexity Ai and whoever else dreams

play11:47

something up when it comes to search in

play11:50

their mom's

play11:53

garage I am really excited to see what

play11:56

comes of this and this is where I go

play12:00

we need to be looking for the small

play12:03

company startups small cap and some

play12:07

cases micro cap some cases we can't

play12:09

invest in them and find out what

play12:12

companies are really interested in them

play12:14

because those are the companies who will

play12:16

eventually acquire them in

play12:20

Amazon being one of those players SK

play12:23

Telecom being one of those players Nidia

play12:27

being one of those players who is taking

play12:31

notice and putting money behind these

play12:34

guys to help them develop the next level

play12:37

of generative Ai and

play12:40

RL and

play12:43

search so keep an eye on this I'll leave

play12:45

a link to their website provy ai's

play12:48

website in below and of course this you

play12:52

can learn so much more if you subscribe

play12:55

to the Platinum Channel at best of us

play12:58

investors we got a 50% off 3 months for

play13:01

3 months uh program off the $60 a

play13:04

monthly fee you get a whole bunch of

play13:06

stuff go to our website check it out but

play13:09

most

play13:10

importantly whatever you

play13:12

do ask questions be a better question

play13:16

asker and be curious because this is

play13:20

where we find the Nuggets this is where

play13:23

we find the

play13:25

eventual line of gold through the side

play13:28

of the mountain

play13:29

live

play13:32

[Music]

play13:38

loud

Rate This

5.0 / 5 (0 votes)

Related Tags
AI SearchPerplexity AIChatGPTGoogle DisruptionConversational AINatural LanguageTech InnovationInvestment OpportunitiesFuturistic SearchSearch Monopoly