June 19, 2024 • 21 mins
Ever wondered if AI chat interfaces could revolutionize your search experience? Imagine having a virtual assistant that not only answers your queries but understands the context and nuances behind them. In this episode, Philip Mastroianni takes you on a journey through the world of AI-powered chatbots like ChatGPT, shedding light on their growing popularity and transformative impact on traditional search engines like Google. We’ll discuss Perplexity AI, a unique search engine that combines conversational abilities with real-time data retrieval, offering a more personalized and interactive approach to finding information. Philip compares Perplexity's strengths and limitations against traditional search engines, providing a comprehensive look at how these advanced tools are changing the landscape of online searches.
Looking to the future, Philip explores the emerging trend of RAG custom chats and their potential to provide even more personalized and nuanced search experiences. While these specialized chats require substantial upfront effort to curate their knowledge base, the payoff can be significant for users within specific knowledge domains. The episode also tackles the critical issues of trust and credibility in AI-generated content, highlighting the necessary balance between innovative AI tools and traditional search engines. This conversation delves into how AI technology might revolutionize SEO and information retrieval while emphasizing the continued importance of a diverse ecosystem of search tools tailored to different user needs and preferences. Don’t miss this insightful discussion on the future of search technology!
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
Welcome to the Opinionated SEO.
I'm Philip Mastroianni, andtoday I want to talk about
something that's been on my mindlately and that's really
looking at the rise of theseAI-powered chat bots and
interfaces and how things havebeen shifting in the way we're
searching online, and thatreally got me thinking.
What does this really mean forsearch engines like Google?
(00:22):
As someone who works in SEO,I'm always curious about how
things might change the way weapproach search in general.
In this episode, I'm going toexplore the idea that AI chats
are going to erode the marketshare of traditional search
engines, and not just by a fewpercentage points, but I mean
like 10-20%.
(00:43):
We're going to take a look atan example of an AI search
engine Perplexity AI and discussits strengths and weaknesses,
compare it to some othertraditional search engines like
Google and also RAG interfaces,and how they step up.
What's behind this potentialshift in search behavior?
Well, it's all about AI chats.
(01:03):
You've probably heard of them,right?
Chatgpt You've probably usedthem yourself.
There are these conversationalinterfaces that can understand
natural language and actuallyrespond with helpful answers.
Now AI chats are becomingincreasingly popular.
It's really not hard to see whythey offer a more personalized
(01:23):
and interactive experiencecompared to traditional search
engines.
You can ask follow-up questions, get more detailed explanations
.
You can even have aconversation that feels a lot
more human-like.
But what really sets theseapart is their ability to be
trained on specific data sets.
This means you can have ahighly tailored discussion
(01:43):
that's focused on a particulartopic or industry, and the best
part is, you can refine yourquestions and get more precise
answers without having to siftthrough a ton of irrelevant
search results.
It's like having a personalexpert or researcher at your
fingertips and they're providingexactly the information you
need.
I saw a great example of a chatthat was trained on a very
(02:06):
specific professional camera'smanual and the person was asking
the chatbot questions thatdidn't quite align with the
right terminology.
The chatbot understood thesimilarity with the topic and
knew the information that theywere actually looking for, and
was able to retrieve and answertheir question and help them
understand how to achieve whatthey were looking for, even
(02:28):
though they didn't know what thefeature was called.
This is like having a personalassistant that can help you find
exactly what you're looking forwithout having to dig through a
bunch of irrelevant searchresults.
And that's what's so excitingabout AI chats they have the
potential to revolutionize theway we search for information
online.
Instead of relying on broadsearch queries, we can have more
(02:50):
focused and targetedconversations that get us
exactly what we need With therise of these powerful AI chat
interfaces, it got me thinking.
What does this mean for thefuture of traditional search
engines like Google?
One platform that's trying toblend the best of both worlds is
Perplexity AI.
Perplexity is an AI-poweredsearch engine designed to
(03:11):
provide a more conversationaland interactive search
experience.
So, unlike a typical Googlesearch, perplexity aims to
understand the context andnuance much more behind your
queries and deliver veryspecific, tailored responses
that feel much more like talkingto a knowledgeable expert.
One of the other things thatthey have that's outside of many
(03:32):
other types of chatbots is theypull real-time data.
So let's take a little bit of acloser look at Perplexity what
they have to offer and maybe howit compares to the search
experience that we're morefamiliar with on Google.
Perplexity is an AI-poweredsearch engine that aims to
provide a more conversationaland interactive search
(03:53):
experience.
But what does that actuallylook like in practice?
One of Perplexity's keystrengths is its ability to
excel at in-depth research andanalysis, unlike a traditional
search engine that may justreturn a list of links.
Perplexity can provide answersby synthesizing that information
from multiple sources in realtime.
(04:14):
So, for example, let's say youneed to do some thorough
research on a complex topic likeenvironmental impact of
cryptocurrency mining.
With Perplexity, you could ask adetailed question and it would
respond with a well-structuredsummary, citing relevant data
and studies, to give you aholistic understanding of the
issue.
It can help you understand itbetter by teaching you what
(04:37):
certain terms are, as well asgiving you the information in
different ways.
It allows you to keep followingup with those questions and
asking to help change theresponse so that you can
understand it easier.
And this is how the tool shinesin these types of
conversational queries Itsnatural language processing
abilities allow you to askfollow-up questions and engage
(05:00):
in a back and forth dialogue soyou can really fully explore a
topic.
And this is great for thosetimes when you have a lot of
nuanced questions on your mindand want a more guided research
experience, or don't quiteunderstand the topic and need
additional guidance to reallytruly understand and to dig in,
so to give an idea of how asearch might be performed that
(05:21):
ends up tailoring to yourspecific needs for a real world
scenario.
Let's say you have an upcomingjob interview for a senior
marketing manager role at a techcompany.
You could ask perplexitysomething like what are the most
important things I should knowand prepare for in a job
interview for a senior marketingmanager position at a tech
company?
I'm currently a senior managerand looking to move to a
(05:44):
different company.
Well, rather than justreturning a generic list of
interview tips, perplexity wouldtry to understand the specific
context of your situation theseniority level, the industry,
your own background.
Context of your situation, theseniority level, the industry,
your own background.
The content is much different,based on the context of your
question, and it allows forfollow-ups that continue to
adhere to that context.
Now, perplexity is not withoutits limitations.
(06:07):
It's an AI-powered system, itcan sometimes have biases, blind
spots, and it's knowledge-based.
And though it does pull from awide range of sources, there
could be certain niche orspecific topics where its
coverage is just not ascomprehensive or it doesn't have
that information.
It doesn't know where to searchto find it.
So, while perplexity representslike a really interesting
(06:29):
evolution in search, it'simportant to be aware of both
its strengths and weaknesses andwhen it might be the optimal
tool compared to a moretraditional approach like a
Google search.
Now you're also getting mostlysummarized data from third-party
sources and you may want tofact check with the citations
given.
Now, in many cases, this isn'tan issue, but in more critical
(06:51):
areas it may be better to justgo straight to the source.
So if you're looking to do aquick factual lookup finding
business hours, addresses,weather forecasts Google search
results may be morestraightforward and immediately
useful.
Perplexity excels at providingmore of the nuanced, synthesized
information, but sometimes youjust need a simple, direct
(07:12):
answer found from a website.
Google search also tends to bebetter suited for things like
localized search, finding nearbyrestaurants, shops or services
and because of its deepintegration with maps, reviews
and other local data, it reallymakes it the go-to for those
kinds of location-based searchesand for certain tasks like
(07:32):
online shopping or travelplanning, google search results
often are more directlyactionable.
It's tight connections toe-commerce platforms and travel
booking sites can make it moreconvenient for those types of
use cases.
If you're not sure where youwant to travel to, perplexity
may be a great way to startasking those kinds of questions
(07:53):
and narrowing things down, butactually booking your flight and
coming up with some of thosenuanced components to your
search, google may be the way togo.
Sometimes you're just trying toaccess a specific website or
online resource, and Google canusually get you there much more
efficiently than an AI chatinterface like Perplexity.
The direct search results andwebsite links are often the
(08:15):
quickest path.
Especially if you don't knowthe right domain for a brand or
restaurant, google may be thefastest solution.
Perplexity shines for in-depthresearch analysis,
conversational queries where youneed a more tailored, nuanced
response, but Google remains theoptimal choice for quick
factual lookups, localizedsearch, e-commerce and directly
(08:36):
accessing websites.
Now, understanding thestrengths and limitations of
each approach really is going tohelp you determine the right
type of tool for the job.
And while perplexity and Googlesearch each have their own
strengths and weaknesses,there's another category of
AI-powered interfaces that areworth exploring, and this is the
general knowledge chatbot.
(08:57):
So you've got platforms likeChatGPT, anthropics, cloud.
They've really captured just aton of attention for their
ability to engage infree-flowing conversation on a
wide range of topics, unlikesearch engines that result lists
of results or more specializedAI assistants like Perplexity,
which work to summarize thoseresults.
Generally, these generalknowledge chatbots are designed
(09:18):
to be knowledgeable companionsthat can help you explore ideas,
answer questions and eventackle complex tasks.
One thing to note is that thereare open source models that can
be run locally on your computeras well, and so you could have
these types of conversationswith these general knowledge
chatbots without even havinginternet.
(09:41):
The key differentiator as wellis the breadth of knowledge base
.
Rather than being trained on aspecific domain or dataset,
these chatbots have been exposedto a massive amount of
information spanning science,history, some current events,
past events, creative writing,history you name it.
This allows them to converse onjust about any subject, and
(10:04):
they can draw insights and makeconnections that a traditional
search just might miss.
But it's important to note that, while these general chatbots
have impressive depths ofknowledge, that information is
not nearly or necessarily asup-to-date as something like
Perplexity's real-time datapulled directly from the web.
The free versions of thesechatbots, like your chat GPTs in
(10:25):
particular, may have knowledgebases that can become stale over
time, especially when it comesto rapidly evolving current
events or time-sensitiveinformation.
If you're running a local model, it is literally when that
model is created is the lastcutoff, so it doesn't have any
new information that it can addto it.
That said, these generalknowledge.
(10:47):
Chatbots aren't without theirstrengths.
Their ability to converse on avast array of subjects, making
connections and providingnuanced insights can be
incredibly valuable foropen-ended exploratory queries
where you're looking to divedeep into a topic.
When weighing the differentAI-powered search and
conversational options, it'simportant to understand the
(11:11):
unique strengths and trade-offsof each approach.
Perplexity, google search andgeneral chatbots all have their
place in the evolving landscapeof how we find and engage with
information.
We've explored the capabilitiesof perplexity, which provides a
more specialized data-drivensearch experience, as well as
the broad knowledge andconversational abilities of
(11:31):
general AI chatbots.
But there's another interestingcategory of AI-powered
interfaces worth discussing RAGcustom chats.
Rag stands for Retrieval,augmented Generation, and it
refers to the type of AI systemthat can take a specific piece
of knowledge or information andthen engage in a customized
question and answer dialoguearound that content.
(11:52):
For example, let's say there'sa detailed industry report on
the impact of artificialintelligence on business
strategy and operation.
With RAG Custom Chat Interface,you could upload that report
and then ask the AI a series offollow-up questions, everything
from high-level summaries todrilling down into specific data
points, emerging trends andrecommended approaches.
The AI would leverage theinformation in the report to
(12:16):
provide answers, clarificationsand even suggest additional
areas for exploration.
It could walk you through keyfindings, help you interpret the
data and even brainstorm howthe insights from the report
could be applied to your ownbusiness challenges.
The key strength of RAG customchats is their ability to
provide a truly personalizedinteractive experience.
(12:36):
Unlike a general search engineor chatbot, which may struggle
to give comprehensive orcontextual responses, these AI
assistants are laser focused ona specific knowledge domain.
This allows them to have a morenuanced back and forth
discussion, really helps usersget the most out of the
available information.
However, rag custom chatsaren't without limitation, so
(12:59):
for one, they require theupfront work of actually
curating and getting thatknowledge base that the AI is
going to draw from.
This can be possibly timeconsuming and could limit the
breadth of topics availablecompared to a more open-ended
chatbot.
Additionally, the quality ofthe response is still dependent
on the depth and accuracy of theunderlying data.
(13:20):
So bad data in, bad data out,and it's really important to
note that the more quality datayou put in there, the more that
it's going to have to work with.
But when you start askingquestions outside of that data
set, it's going to have to fallback onto its just general
knowledge base, and so it'sreally important to make sure
that that stays focused.
(13:40):
Now, the RAC systems can providevery tailored insights, but
they aren't immune to biases orgaps in their knowledge base.
So when you're comparing thedifferent RAC custom chats to
other AI-powered tools that wediscussed, the key distinction
is really that level of focusand customization.
Perplexity excels at providingwide-ranging, up-to-date
(14:02):
information.
General chatbots are great atopen-ended exploration.
Rag custom chats, on the otherhand, aim to deliver a highly
curated interactive experiencecentered around a unique,
specific knowledge domain.
So, ultimately, each approachhas its own strengths and
weaknesses, and the optimalchoice will really depend on the
(14:22):
user's specific needs and thetype of information they're
seeking.
Understanding the uniquecapabilities of RAC, perplexity,
general chatbots and Google canhelp you determine the right
tool for the job.
We covered a lot of ground hereAI chatbots, perplexity chat,
gpt, google search, how they allkind of have their own nuanced
(14:43):
ways of searching, even thespecialized capabilities of RAG
custom chats.
It's really that dive into theevolving world of information,
discovery and search.
But the big question is whatdoes all of this mean for the
future, and especially thefuture of Google search?
So let's be real Traditionalsearch engines like Google have
pretty much dominated the gamefor the past couple of decades
(15:05):
and they've certainly tried tostay ahead of the curve, rolling
out their own AI-powered searchfeatures like SGE and, more
recently, ai overviews.
Bing has partnered up withOpenAI and they've got their own
going, but the reception tothese efforts has really been
lukewarm at best.
A big part of the challenge isthat these AI-infused search
results don't quite capture thenuanced and conversational flow
(15:28):
of a true chatbot experience.
The responses can feel a bitstilted and disconnected,
lacking that fluidity andcontextual understanding that
makes platforms like Perplexity,a RAG, custom chat or your just
general chat GPTs just socompelling.
There's also the issue of trustand credibility.
(15:49):
Google's search results havelong been seen as a gold
standard for authoritative,up-to-date information, but when
you start injecting more AIgenerated content into the mix,
users can start to question thereliability and accuracy of what
they're seeing.
Remember, google presented uswith what they felt was the best
(16:09):
result and we made the choiceof which ones we were going to
look at.
The AI powering these searchfeatures, while impressive the
technology is incrediblyimpressive they still exhibit
biases.
They have gaps in theirknowledge.
They lack the nuance thattraditional search engines have
largely been able to avoid, andusers may be wary of relying on
(16:33):
AI curated results, especiallyfor important decisions or those
high-stake queries.
I don't foresee Google evergoing away for those types of
requests, so the idea ofblending the best of traditional
search and conversational AI isenticing.
The execution has sucked.
No one has been impressed withit so far.
(16:55):
The user experience justdoesn't feel as seamless or
trustworthy as a truly dedicatedAI interface or a
well-established search engine.
At the end of the day, I thinkthe key is having a diverse
ecosystem of search anddiscovery tools, each catering
to different needs andpreferences.
Ai chatbots may never fullyreplace traditional search, but
(17:16):
they could become powerfulcompanions, really enhancing our
ability to find, understand andapply information in more
personalized, efficient ways.
The search landscape's evolving.
It's going to be fascinating tosee how Google and other
players adapt, but for now, themelding of these two approaches
doesn't quite seem to beresonating with users in the way
that the tech giants had likelyhoped.
(17:38):
It's a challenge they'll needto continue refining and
perfecting.
Well, there you have it.
That's my dive into theevolving world of AI-powered
search and discovery, fromconversational chatbots to
specialized capabilities oftools like Perplexity and RAG,
it's clear that are offeringcompelling alternatives that
(17:59):
could start chipping away attheir market share.
The ability to have naturalback and forth dialogues, get
tailored insights and exploreknowledge in more interactive
(18:19):
ways is actually really exciting.
But, of course, each approachhas its own strengths and
weaknesses, and really it'simportant that you understand
the nuances between thedifferent platforms like
perplexity, these generalchatbots, rag interfaces,
because it's going to be crucialto find the right tool for the
job, and as people use thesesystems more and more, they're
going to know where they'regoing to go and more and more
(18:41):
people are going to go away fromGoogle to search in these more
conversational ways, because thefuture of search isn't blending
all of these together, but it'sreally making them the best for
the type of search that youhave.
So, personally, I think it's abig mistake for Google to double
down on AI overviews in theirsearch results.
The data they're pulling isn'talways vetted.
(19:02):
Anyone can post their opinionor even satirical content that
may look accurate.
The kind of unfilteredinformation can be dangerous,
especially for queries aroundimportant, time-sensitive topics
, and it's the kind of thingsthat we've trained ourselves to
spot, but these AI overviewsystems just haven't been able
to figure it out.
(19:22):
So generalized chatbots, on theother hand, great for getting
custom information on lesstime-critical subjects, and
specialized AI search engineslike Perplexity have a real
place where you want morenuanced responses that leverage
real-time data.
In the end, I believe Google isgoing to lose double-digit
market share as people migrateaway from them for many types of
(19:43):
queries.
The appeal of havingconversational contextual search
experience is just too strongand in the future, because of
the technology advances, theselocal models, like Apple's new
AI, is going to run locally onyour phone and not even require
internet access to run.
I wouldn't be surprised to seemore and more users creating
(20:04):
their own custom chatbots,trained on their emails, company
wikis or personal knowledgebase, or just built into things
like their phone, where it hasall of their data already and
you could ask it questions, andit's going to have that as its
data source.
The future of search isexciting, but it's also a bit
unnerving, especially if you'rean SEO.
So we need to be vigilant aboutthe quality and integrity of
(20:25):
the information we're relying on.
Really, make sure you'relooking at those citations, but
the potential of theseAI-powered tools to
revolutionize how we find,understand and apply knowledge
is pretty undeniable.
So what do you think?
Are you as bullish as I am onthe rise of AI, chats and custom
search experience, or do youthink the old school search
engine will remain king?
I'd love to hear your thoughtsand opinions Reach out to me.
(20:46):
Hello at opinionatedseocom.
Let me know In the meantime,stay tuned for some more future
episodes.
I got a couple people lined upwe're going to be interviewing,
who are going to talk about SEOand AI and some other things
that are really exciting, andwe'll unpack it all and see what
that means for digitalmarketing and SEO.
Welcome to the Opinionated SEO.
I'm Philip Mastroianni, andtoday I want to talk about
something that's been on my mindlately and that's really
looking at the rise of theseAI-powered chat bots and
interfaces and how things havebeen shifting in the way we're
searching online, and thatreally got me thinking.
What does this really mean forsearch engines like Google?
(00:22):
As someone who works in SEO,I'm always curious about how
things might change the way weapproach search in general.
In this episode, I'm going toexplore the idea that AI chats
are going to erode the marketshare of traditional search
engines, and not just by a fewpercentage points, but I mean
like 10-20%.
(00:43):
We're going to take a look atan example of an AI search
engine Perplexity AI and discussits strengths and weaknesses,
compare it to some othertraditional search engines like
Google and also RAG interfaces,and how they step up.
What's behind this potentialshift in search behavior?
Well, it's all about AI chats.
(01:03):
You've probably heard of them,right?
Chatgpt You've probably usedthem yourself.
There are these conversationalinterfaces that can understand
natural language and actuallyrespond with helpful answers.
Now AI chats are becomingincreasingly popular.
It's really not hard to see whythey offer a more personalized
(01:23):
and interactive experiencecompared to traditional search
engines.
You can ask follow-up questions, get more detailed explanations
.
You can even have aconversation that feels a lot
more human-like.
But what really sets theseapart is their ability to be
trained on specific data sets.
This means you can have ahighly tailored discussion
(01:43):
that's focused on a particulartopic or industry, and the best
part is, you can refine yourquestions and get more precise
answers without having to siftthrough a ton of irrelevant
search results.
It's like having a personalexpert or researcher at your
fingertips and they're providingexactly the information you
need.
I saw a great example of a chatthat was trained on a very
(02:06):
specific professional camera'smanual and the person was asking
the chatbot questions thatdidn't quite align with the
right terminology.
The chatbot understood thesimilarity with the topic and
knew the information that theywere actually looking for, and
was able to retrieve and answertheir question and help them
understand how to achieve whatthey were looking for, even
(02:28):
though they didn't know what thefeature was called.
This is like having a personalassistant that can help you find
exactly what you're looking forwithout having to dig through a
bunch of irrelevant searchresults.
And that's what's so excitingabout AI chats they have the
potential to revolutionize theway we search for information
online.
Instead of relying on broadsearch queries, we can have more
(02:50):
focused and targetedconversations that get us
exactly what we need With therise of these powerful AI chat
interfaces, it got me thinking.
What does this mean for thefuture of traditional search
engines like Google?
One platform that's trying toblend the best of both worlds is
Perplexity AI.
Perplexity is an AI-poweredsearch engine designed to
(03:11):
provide a more conversationaland interactive search
experience.
So, unlike a typical Googlesearch, perplexity aims to
understand the context andnuance much more behind your
queries and deliver veryspecific, tailored responses
that feel much more like talkingto a knowledgeable expert.
One of the other things thatthey have that's outside of many
(03:32):
other types of chatbots is theypull real-time data.
So let's take a little bit of acloser look at Perplexity what
they have to offer and maybe howit compares to the search
experience that we're morefamiliar with on Google.
Perplexity is an AI-poweredsearch engine that aims to
provide a more conversationaland interactive search
(03:53):
experience.
But what does that actuallylook like in practice?
One of Perplexity's keystrengths is its ability to
excel at in-depth research andanalysis, unlike a traditional
search engine that may justreturn a list of links.
Perplexity can provide answersby synthesizing that information
from multiple sources in realtime.
(04:14):
So, for example, let's say youneed to do some thorough
research on a complex topic likeenvironmental impact of
cryptocurrency mining.
With Perplexity, you could ask adetailed question and it would
respond with a well-structuredsummary, citing relevant data
and studies, to give you aholistic understanding of the
issue.
It can help you understand itbetter by teaching you what
(04:37):
certain terms are, as well asgiving you the information in
different ways.
It allows you to keep followingup with those questions and
asking to help change theresponse so that you can
understand it easier.
And this is how the tool shinesin these types of
conversational queries Itsnatural language processing
abilities allow you to askfollow-up questions and engage
(05:00):
in a back and forth dialogue soyou can really fully explore a
topic.
And this is great for thosetimes when you have a lot of
nuanced questions on your mindand want a more guided research
experience, or don't quiteunderstand the topic and need
additional guidance to reallytruly understand and to dig in,
so to give an idea of how asearch might be performed that
(05:21):
ends up tailoring to yourspecific needs for a real world
scenario.
Let's say you have an upcomingjob interview for a senior
marketing manager role at a techcompany.
You could ask perplexitysomething like what are the most
important things I should knowand prepare for in a job
interview for a senior marketingmanager position at a tech
company?
I'm currently a senior managerand looking to move to a
(05:44):
different company.
Well, rather than justreturning a generic list of
interview tips, perplexity wouldtry to understand the specific
context of your situation theseniority level, the industry,
your own background.
Context of your situation, theseniority level, the industry,
your own background.
The content is much different,based on the context of your
question, and it allows forfollow-ups that continue to
adhere to that context.
Now, perplexity is not withoutits limitations.
(06:07):
It's an AI-powered system, itcan sometimes have biases, blind
spots, and it's knowledge-based.
And though it does pull from awide range of sources, there
could be certain niche orspecific topics where its
coverage is just not ascomprehensive or it doesn't have
that information.
It doesn't know where to searchto find it.
So, while perplexity representslike a really interesting
(06:29):
evolution in search, it'simportant to be aware of both
its strengths and weaknesses andwhen it might be the optimal
tool compared to a moretraditional approach like a
Google search.
Now you're also getting mostlysummarized data from third-party
sources and you may want tofact check with the citations
given.
Now, in many cases, this isn'tan issue, but in more critical
(06:51):
areas it may be better to justgo straight to the source.
So if you're looking to do aquick factual lookup finding
business hours, addresses,weather forecasts Google search
results may be morestraightforward and immediately
useful.
Perplexity excels at providingmore of the nuanced, synthesized
information, but sometimes youjust need a simple, direct
(07:12):
answer found from a website.
Google search also tends to bebetter suited for things like
localized search, finding nearbyrestaurants, shops or services
and because of its deepintegration with maps, reviews
and other local data, it reallymakes it the go-to for those
kinds of location-based searchesand for certain tasks like
(07:32):
online shopping or travelplanning, google search results
often are more directlyactionable.
It's tight connections toe-commerce platforms and travel
booking sites can make it moreconvenient for those types of
use cases.
If you're not sure where youwant to travel to, perplexity
may be a great way to startasking those kinds of questions
(07:53):
and narrowing things down, butactually booking your flight and
coming up with some of thosenuanced components to your
search, google may be the way togo.
Sometimes you're just trying toaccess a specific website or
online resource, and Google canusually get you there much more
efficiently than an AI chatinterface like Perplexity.
The direct search results andwebsite links are often the
(08:15):
quickest path.
Especially if you don't knowthe right domain for a brand or
restaurant, google may be thefastest solution.
Perplexity shines for in-depthresearch analysis,
conversational queries where youneed a more tailored, nuanced
response, but Google remains theoptimal choice for quick
factual lookups, localizedsearch, e-commerce and directly
(08:36):
accessing websites.
Now, understanding thestrengths and limitations of
each approach really is going tohelp you determine the right
type of tool for the job.
And while perplexity and Googlesearch each have their own
strengths and weaknesses,there's another category of
AI-powered interfaces that areworth exploring, and this is the
general knowledge chatbot.
(08:57):
So you've got platforms likeChatGPT, anthropics, cloud.
They've really captured just aton of attention for their
ability to engage infree-flowing conversation on a
wide range of topics, unlikesearch engines that result lists
of results or more specializedAI assistants like Perplexity,
which work to summarize thoseresults.
Generally, these generalknowledge chatbots are designed
(09:18):
to be knowledgeable companionsthat can help you explore ideas,
answer questions and eventackle complex tasks.
One thing to note is that thereare open source models that can
be run locally on your computeras well, and so you could have
these types of conversationswith these general knowledge
chatbots without even havinginternet.
(09:41):
The key differentiator as wellis the breadth of knowledge base
.
Rather than being trained on aspecific domain or dataset,
these chatbots have been exposedto a massive amount of
information spanning science,history, some current events,
past events, creative writing,history you name it.
This allows them to converse onjust about any subject, and
(10:04):
they can draw insights and makeconnections that a traditional
search just might miss.
But it's important to note that, while these general chatbots
have impressive depths ofknowledge, that information is
not nearly or necessarily asup-to-date as something like
Perplexity's real-time datapulled directly from the web.
The free versions of thesechatbots, like your chat GPTs in
(10:25):
particular, may have knowledgebases that can become stale over
time, especially when it comesto rapidly evolving current
events or time-sensitiveinformation.
If you're running a local model, it is literally when that
model is created is the lastcutoff, so it doesn't have any
new information that it can addto it.
That said, these generalknowledge.
(10:47):
Chatbots aren't without theirstrengths.
Their ability to converse on avast array of subjects, making
connections and providingnuanced insights can be
incredibly valuable foropen-ended exploratory queries
where you're looking to divedeep into a topic.
When weighing the differentAI-powered search and
conversational options, it'simportant to understand the
(11:11):
unique strengths and trade-offsof each approach.
Perplexity, google search andgeneral chatbots all have their
place in the evolving landscapeof how we find and engage with
information.
We've explored the capabilitiesof perplexity, which provides a
more specialized data-drivensearch experience, as well as
the broad knowledge andconversational abilities of
(11:31):
general AI chatbots.
But there's another interestingcategory of AI-powered
interfaces worth discussing RAGcustom chats.
Rag stands for Retrieval,augmented Generation, and it
refers to the type of AI systemthat can take a specific piece
of knowledge or information andthen engage in a customized
question and answer dialoguearound that content.
(11:52):
For example, let's say there'sa detailed industry report on
the impact of artificialintelligence on business
strategy and operation.
With RAG Custom Chat Interface,you could upload that report
and then ask the AI a series offollow-up questions, everything
from high-level summaries todrilling down into specific data
points, emerging trends andrecommended approaches.
The AI would leverage theinformation in the report to
(12:16):
provide answers, clarificationsand even suggest additional
areas for exploration.
It could walk you through keyfindings, help you interpret the
data and even brainstorm howthe insights from the report
could be applied to your ownbusiness challenges.
The key strength of RAG customchats is their ability to
provide a truly personalizedinteractive experience.
(12:36):
Unlike a general search engineor chatbot, which may struggle
to give comprehensive orcontextual responses, these AI
assistants are laser focused ona specific knowledge domain.
This allows them to have a morenuanced back and forth
discussion, really helps usersget the most out of the
available information.
However, rag custom chatsaren't without limitation, so
(12:59):
for one, they require theupfront work of actually
curating and getting thatknowledge base that the AI is
going to draw from.
This can be possibly timeconsuming and could limit the
breadth of topics availablecompared to a more open-ended
chatbot.
Additionally, the quality ofthe response is still dependent
on the depth and accuracy of theunderlying data.
(13:20):
So bad data in, bad data out,and it's really important to
note that the more quality datayou put in there, the more that
it's going to have to work with.
But when you start askingquestions outside of that data
set, it's going to have to fallback onto its just general
knowledge base, and so it'sreally important to make sure
that that stays focused.
(13:40):
Now, the RAC systems can providevery tailored insights, but
they aren't immune to biases orgaps in their knowledge base.
So when you're comparing thedifferent RAC custom chats to
other AI-powered tools that wediscussed, the key distinction
is really that level of focusand customization.
Perplexity excels at providingwide-ranging, up-to-date
(14:02):
information.
General chatbots are great atopen-ended exploration.
Rag custom chats, on the otherhand, aim to deliver a highly
curated interactive experiencecentered around a unique,
specific knowledge domain.
So, ultimately, each approachhas its own strengths and
weaknesses, and the optimalchoice will really depend on the
(14:22):
user's specific needs and thetype of information they're
seeking.
Understanding the uniquecapabilities of RAC, perplexity,
general chatbots and Google canhelp you determine the right
tool for the job.
We covered a lot of ground hereAI chatbots, perplexity chat,
gpt, google search, how they allkind of have their own nuanced
(14:43):
ways of searching, even thespecialized capabilities of RAG
custom chats.
It's really that dive into theevolving world of information,
discovery and search.
But the big question is whatdoes all of this mean for the
future, and especially thefuture of Google search?
So let's be real Traditionalsearch engines like Google have
pretty much dominated the gamefor the past couple of decades
(15:05):
and they've certainly tried tostay ahead of the curve, rolling
out their own AI-powered searchfeatures like SGE and, more
recently, ai overviews.
Bing has partnered up withOpenAI and they've got their own
going, but the reception tothese efforts has really been
lukewarm at best.
A big part of the challenge isthat these AI-infused search
results don't quite capture thenuanced and conversational flow
(15:28):
of a true chatbot experience.
The responses can feel a bitstilted and disconnected,
lacking that fluidity andcontextual understanding that
makes platforms like Perplexity,a RAG, custom chat or your just
general chat GPTs just socompelling.
There's also the issue of trustand credibility.
(15:49):
Google's search results havelong been seen as a gold
standard for authoritative,up-to-date information, but when
you start injecting more AIgenerated content into the mix,
users can start to question thereliability and accuracy of what
they're seeing.
Remember, google presented uswith what they felt was the best
(16:09):
result and we made the choiceof which ones we were going to
look at.
The AI powering these searchfeatures, while impressive the
technology is incrediblyimpressive they still exhibit
biases.
They have gaps in theirknowledge.
They lack the nuance thattraditional search engines have
largely been able to avoid, andusers may be wary of relying on
(16:33):
AI curated results, especiallyfor important decisions or those
high-stake queries.
I don't foresee Google evergoing away for those types of
requests, so the idea ofblending the best of traditional
search and conversational AI isenticing.
The execution has sucked.
No one has been impressed withit so far.
(16:55):
The user experience justdoesn't feel as seamless or
trustworthy as a truly dedicatedAI interface or a
well-established search engine.
At the end of the day, I thinkthe key is having a diverse
ecosystem of search anddiscovery tools, each catering
to different needs andpreferences.
Ai chatbots may never fullyreplace traditional search, but
(17:16):
they could become powerfulcompanions, really enhancing our
ability to find, understand andapply information in more
personalized, efficient ways.
The search landscape's evolving.
It's going to be fascinating tosee how Google and other
players adapt, but for now, themelding of these two approaches
doesn't quite seem to beresonating with users in the way
that the tech giants had likelyhoped.
(17:38):
It's a challenge they'll needto continue refining and
perfecting.
Well, there you have it.
That's my dive into theevolving world of AI-powered
search and discovery, fromconversational chatbots to
specialized capabilities oftools like Perplexity and RAG,
it's clear that are offeringcompelling alternatives that
(17:59):
could start chipping away attheir market share.
The ability to have naturalback and forth dialogues, get
tailored insights and exploreknowledge in more interactive
(18:19):
ways is actually really exciting.
But, of course, each approachhas its own strengths and
weaknesses, and really it'simportant that you understand
the nuances between thedifferent platforms like
perplexity, these generalchatbots, rag interfaces,
because it's going to be crucialto find the right tool for the
job, and as people use thesesystems more and more, they're
going to know where they'regoing to go and more and more
(18:41):
people are going to go away fromGoogle to search in these more
conversational ways, because thefuture of search isn't blending
all of these together, but it'sreally making them the best for
the type of search that youhave.
So, personally, I think it's abig mistake for Google to double
down on AI overviews in theirsearch results.
The data they're pulling isn'talways vetted.
(19:02):
Anyone can post their opinionor even satirical content that
may look accurate.
The kind of unfilteredinformation can be dangerous,
especially for queries aroundimportant, time-sensitive topics
, and it's the kind of thingsthat we've trained ourselves to
spot, but these AI overviewsystems just haven't been able
to figure it out.
(19:22):
So generalized chatbots, on theother hand, great for getting
custom information on lesstime-critical subjects, and
specialized AI search engineslike Perplexity have a real
place where you want morenuanced responses that leverage
real-time data.
In the end, I believe Google isgoing to lose double-digit
market share as people migrateaway from them for many types of
(19:43):
queries.
The appeal of havingconversational contextual search
experience is just too strongand in the future, because of
the technology advances, theselocal models, like Apple's new
AI, is going to run locally onyour phone and not even require
internet access to run.
I wouldn't be surprised to seemore and more users creating
(20:04):
their own custom chatbots,trained on their emails, company
wikis or personal knowledgebase, or just built into things
like their phone, where it hasall of their data already and
you could ask it questions, andit's going to have that as its
data source.
The future of search isexciting, but it's also a bit
unnerving, especially if you'rean SEO.
So we need to be vigilant aboutthe quality and integrity of
(20:25):
the information we're relying on.
Really, make sure you'relooking at those citations, but
the potential of theseAI-powered tools to
revolutionize how we find,understand and apply knowledge
is pretty undeniable.
So what do you think?
Are you as bullish as I am onthe rise of AI, chats and custom
search experience, or do youthink the old school search
engine will remain king?
I'd love to hear your thoughtsand opinions Reach out to me.
(20:46):
Hello at opinionatedseocom.
Let me know In the meantime,stay tuned for some more future
episodes.
I got a couple people lined upwe're going to be interviewing,
who are going to talk about SEOand AI and some other things
that are really exciting, andwe'll unpack it all and see what
that means for digitalmarketing and SEO.
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by audiochuck Media Company.