 |
| Is Google's Reign Over? ChatGPT Emerges As A Serious Competitor Gadgets 36T |
Of
course, there were search engines before him. But Google was the first to
popularize knowledge-based search. And while competitors have been around for
years, most are just variations on a
theme. Nowadays, when we want to find something, no one says "I'll dig it
up" or "I'll look it up on Yahoo". However, could all of that be
about to change? Unless you've spent the last few months on Pluto, you've
probably noticed that ChatGPT is the "hot topic" app at the moment.
Just in case, here's a quick rundown of what it is and how it works. As I sit down to write this article about
whether ChatGPT will finally end
Google's 25-year domination of the search market, it's been reported that
Microsoft has enhanced its Bing search engine by integrating integrates ChatGPT
functionality directly into its interface.
This is not surprising, as Microsoft was one of the first organizations
to support the creators of ChatGPT - the OpenAI research agency - when they
invested $1 billion in 2019. Earlier
this year, they announced that they will continue to invest another $10
billion, making it the largest
shareholder. This clearly brings ChatGPT – and Bing – one step closer to
being the first serious contender for the Google Search throne in a long time.
So should Google be worried? And how does King Research plan to react?
The Data Behind ChatGPT and Google
Search
First, let's see how Google Search and
ChatGPT take different approaches to
solving the same problem - give us the
information we need. Google Search and ChatGPT respond to queries
by searching large databases of information gathered from the Internet.In the
case of ChatGPT, this is the GPT-3 training dataset. The exact contents of this
dataset have not been made public, but it is believed to include 75 billion
parameters (data points), including a 2021 internet scan, the entire content of
Wikipedia, large chunks of Reddit. and two book databases. In total, this
training dataset is reported to be about 45 terabytes in size. This is definitely a large training dataset
compared to other language models. However, it becomes meaningless next to the
data set that Google's search engine uses to answer our queries.Google has been
building its index since the early days of the World Wide Web by sending "crawlers" that move to every
corner of the Internet they can reach. According to Google, the index is
currently around 100,000 terabytes (100 petabytes).
Bigger means better?
However,
as the saying goes, size isn't everything. The main innovation with ChatGPT -
and what makes it, in my opinion, the first serious threat to Google's
dominance in the information economy - is the way it processes this data to
make it useful to us. Basically, Google
returns a list of web pages that, according to Google's algorithm, are
likely to contain the desired information
based on our search queries. On the other hand, ChatGPT uses a Natural
Language Generation (NLG) algorithm to structure the results in a way that
gives us straight answers. It's a huge boost to the user experience. Anyone
using it for research no longer has to wade through search results pages. This experience is very
similar to asking a very knowledgeable friend for their answer and opinion.
However, to be fair, we should mention that Google Search has also included
some AI-powered features – such as the Knowledge Panel – which presents
information extracted from web pages. certain
on the results page. This information is displayed next to or embedded
in the search results of the web pages they return. However, it still cannot
use ChatGPT's conversational style.
This
type of chat is used by ChatGPT to improve user experience. If we do not like
the answer he gives or if we think he
has used the wrong method to help us solve the problems we have presented to
him, we may ask ask him to try again. It will remember previous details about our conversation - at
least until the end of the session - and use this information to refine its
responses until it can give us something
we are satisfied with.
ChatGPT is also ad-free
Most
of the best results we get using Google
search are there because someone paid to
have them there. However, this could
change once ChatGPT hits the market - we'll have to start paying for the
massive amount of computing used when processing our queries at some point and
OpenAI tested "premium" paid versions. several regions, including the
United States and the United Kingdom. Overall, as you would expect from a
technology created 25 years after
Google launched, ChatGPT is light years
ahead in terms of user experience. However, that is far from the end of the
story. Google has a number of advantages that mean it's still far from
over… ChatGPT cons The first issue to deal with is accuracy.
ChatGPT is a brand new tool and is trained on a static, untested dataset.
Unfortunately, this means it's prone to
mistakes and bias - the cause of
much of today's AI technology.With all data processing, the first rule is
"junk equals garbage". Anyone who has been trying to use ChatGPT for
a while may have encountered these inaccuracies. We can now address teething
issues and it's likely his accuracy will improve as he continues to learn and
grow through additional training.
However, this poses another problem.
The fact that ChatGPT
The fact that ChatGPT is somewhat (read: very) unclear about the origin of its data means that fact-checking or verifying its source is difficult. These issues led Google's chief search officer Prabhakar Raghavan to compare the operation of AI-powered chatbots to a "hallucination", describing them as working "in the way that a machine provides something compelling". lead but completely invented". While Google's language processing may seem outdated by comparison, at least Google knows where its information comes from. This is usually the owner and operator of the websites it displays. Of course, this does not mean that everything he tells us will be true and correct - but it will be much simpler to assess the validity and reliability of information Convergence - the future of research ? Of course, these pros and cons are mostly for ChatGPT (and indeed Google Search) as they exist today. In particular, ChatGPT is a nascent technology and what's on offer right now only provides a glimpse of what it - and similar platforms - will be capable of in the near future. One thing's for sure, though, is that Alphabet won't turn around and admit defeat. The commercialization of its search technology has been instrumental in making it one of the richest companies in the world, and it's not a dairy it's giving up any time soon. So, roughly at the same time as Microsoft announced that ChatGPT functionality would be added to Bing, Alphabet said that its own large language model, called Bard, would be used to improve performance. of Google Search.However, things didn't get off to the best start - with machine-made errors in promotional videos blamed for a $100 billion drop in the company's value. But if we were generous enough to attribute the ChatGPT error to a "startup problem", then we would doubt Google's similar interests.
It
is very likely that we will see two technologies - large chat interfaces based on language models and
search engines - combine to create hybrid technologies that will hopefully give
us the best of both worlds. the best of both worlds. Google and Microsoft clearly believe that
the next phase in the evolution of information technology will focus on this convergence
of search and language. And both understand that the catalyst for that will be
AI. Yet another sign that the age of "thinking" machines has been
transforming all aspects of society in ways that just a few years ago seemed unthinkable.
