Google’s Bard and OpenAI’s ChatGPT: Behind the Arms Race

It seems as though the headlines have been dominated by AI recently. From passing bar exam scores to helping the largest retailers in the world fight inventory shrink, artificial intelligence has been a hot topic in every industry. 

Artificial intelligence, some forms of which use NLP, or natural language processing, to create human-like outputs, may stand to threaten search engines the way we know them. 

Google is not about to give up its crown without a fight, so it’s little surprise that they recently unveiled a competitor to ChatGPT, known as Bard. 

Here’s what you need to know about each. 

What Is ChatGPT?

ChatGPT is one of the most capable AI models in the world at the current time. ChatGPT, which stands for Chat Generative Pre-Trained Transformer, is a chatbot developed by OpenAI, an American AI research facility, and released late in November 2022. 

The model predominantly uses two types of machine learning: supervised learning and reinforcement learning, paired with human trainers that fine-tuned and thereby improved the performance of the system. 

The basic function of ChatGPT is to mimic human conversational output based on user input. Consequently, the model has a huge range of different uses, including but not limited to: 

• Writing code/Finding bugs and debugging systems
• Creating apps and scripts 
• Writing resumes and cover letters
• Writing jokes and poetry
• Emulating operating systems 
• Writing novels, web pages, and other creative copy 
• Answering complex queries

How Does ChatGPT Work?

ChatGPT, which has been trained on vast sums of data and fine-tuned by human operators, uses a system called “transformer architecture” to analyze large volumes of data before outputting concise, (relatively) accurate answers to human inputs. 

Since it is open source, effectively anyone can utilize ChatGPT or integrate with the system. The complexity and relevance of the answers it produces are a direct product of the quality of the input. It is possible to create entire websites or articles using a series of complex inputs using ChatGPT’s system.

One important thing to note is that ChatGPT is not connected to the internet, so it is possible that it can create outputs that are either not contemporaneously relevant or which are inaccurate. However, ChatGPT does track user inputs in a thread and enables users to make adjustments and corrections along the way. 

Regardless of its limitations, its relative accuracy and impressive efficiency stand to have great impacts on eCommerce in general and digital marketing in particular. 

ChatGPT, Digital Marketing, and eCommerce

Because ChatGPT has the potential to revolutionize the efficiency of marketing managers, it’s natural to wonder how ChatGPT can possibly impact digital marketing. 

One of the ways in which ChatGPT has the potential to change the game for eCommerce businesses is by improving processes. Artificial intelligence (although not necessarily ChatGPT) has the potential to generate accurate sales and logistics forecasts that improve business processes, reduce lead times, and improve inventory management.

Another area in which artificial intelligence stands to revolutionize eCommerce is by enabling merchants to automate processes that sap human resources. It’s no surprise that a chatbot – like ChatGPT (or models like it) – can potentially improve customer service while reducing the strain on limited human resources. 

AI models like ChatGPT can also streamline the process of keyword research, website coding and development, and content production, all of which are currently central to digital marketing.

It is clear that at the current time (February 2023) AI is neither capable (nor accurate) enough to entirely replace human content writers, ad managers, and developers. With that said, digital marketers are increasingly exploring the potential for AI to supplement their abilities to create conversion and UX-optimized ad copy as well as website and blog content that both meets user search intent and which is optimized for search engine algorithms. 

There is another way in which ChatGPT, or at least a model like it, stands to disrupt the current norms in the world of SEO in particular. 

That is, if human users start using ChatGPT as a search engine, or at least start using it to answer their basic queries instead of using Google. Since ChatGPT uses natural language processing and outputs “natural” human language, it has great potential to answer basic search queries, especially with the rise of voice search. 

With respect to eCommerce (and depending on how these models develop), there is also potential in the future for AI to become the user go-to for searching for online goods and services. 

Adding to this potential has been the vast growth of ChatGPT, which by January 2023 had amassed more than 100 million users, making it the fastest-growing application of all time.

Google has seen the writing on the wall and is not about to let that happen without a fight, so it’s little surprise that only shortly after the release of ChatGPT, Google announced their own AI project: Bard. 

What Is Google Bard?

On February 6th, 2023, Google announced that it too is developing an AI model, which, like ChatGPT, will be trained on a large language model and is a form of generative AI that is expected to produce human-like outputs. 

Though Bard has not been released to the public at the time of this posting, it is expected to have use cases and applications that are similar to ChatGPT. There are, however, a few substantial differences between the two models. 

One is that Bard utilizes a model known as LaMDA, or Language Model for Dialogue Applications, which solves one of the chief issues currently facing chatbots. That is, current chatbots tend to get muddled and produce irrelevant outputs when the discourse between the user and chatbot shifts away from the initial prompt or input.

The LaMDA model, by contrast, is supposedly able to evolve with the natural language of the user – similar to a real-life conversation that evolves as new words and queries are exchanged – enabling a free flow of exchanges that should be (at least in theory) indistinguishable from human conversation. 

Sensibleness is, therefore, a key feature of LaMDA. That is, does the output make sense, given the input. In a natural conversation, occasionally the response to a prompt (comment) is only secondarily or tertiarily relevant. Computers struggle with this, but LaMDA is currently getting better at producing outputs that are relevant (and sensible) to the input.

Another massive potential difference between Bard and ChatGPT is that Bard will not draw from a closed set of data. It will be connected to the internet, which means it will be able to synthesize responses from a larger, more temporally-relevant pool of data. 

Bard’s eventual goal is something that we could expect from Google: to retrieve and display a basic answer to a user input, rather than a search engine results page. The function of both a SERP and an AI output is similar, only the AI output is intended to furnish a more relevant, succinct output. 

Is Google Bard Better Than ChatGPT?

That is a difficult question to answer. The main difference between the two of them is that ChatGPT uses a closed subset of data from 2021 (and earlier), whereas Bard draws from the internet. Consequently, Bard should have access to a larger pool of data (and one that is not only constantly changing but more relevant).

It remains to be seen which model will outperform the other, but neither is free from hiccups. ChatGPT is notorious for producing specious results, and Google’s value recently plummeted by hundred billion in market value after Bard answered a basic question about the James Webb Telescope incorrectly

But it is important to recognize the novelty of these two systems. ChatGPT is only a few months old and Bard is still in beta testing. Their true capabilities (and how they will be used in digital marketing) are yet to materialize, and how they will eventually be used is at this point anyone’s guess. 

Both, however, show great potential at furnishing simple, easy-to-digest answers to basic inputs, which gives them equal potential to be used in the future in a manner similar to how search engines like Google are used today. It is entirely possible that Google intends or merge, or supplement, its current algorithms with advanced artificial intelligence. 

One thing is for sure, though. The investment in artificial intelligence is going nowhere anytime soon, as Google also recently announced that it would be partnering with Anthropic, an AI startup that is being spearheaded by some former members of OpenAI. 

As these nascent additions to the world of computer engineering continue to evolve, it is up to us as digital marketing services providers and eCommerce entrepreneurs to keep abreast of the changes so we can be prepared to implement the best strategies moving forward. 

1Digital^Ⓡ Agency and Your Business: Digital Marketing Strategies for Growth

1Digital^Ⓡ Agency has always been committed to keeping ahead of changes in digital marketing and eCommerce and our team is fully invested in following the most recent changes in the landscape of AI.

Our team is continuously exploring the potential uses of AI in eCommerce, both in terms of how our clients can automate processes and how AI-generated outputs can improve the efficiency and efficacy of our social media management, eCommerce SEO, and PPC campaigns. For more information on the changes in the industry, or how you can leverage AI to the advantage of your eCommerce business’s digital marketing campaigns, get in touch with our digital media specialists at 888-982-8269 or at Sales@1DigitalAgency.com.