OpenAI's New GPTs Trying to Replace Google: Learning Through Conversation

Since the invention of the internet search engine, and Google in particular, "Google it" has become synonymous with learning by reading. This started with web pages, and quickly evolved with the Youtube acquisition to videos: learning by watching instead of reading.

OpenAI and its creation ChatGPT is creating a new paradigm: Learning Through Conversation.

Always the goal of AI, the Turing Test was proposed in 1950 essentially a blind conversation test, that if a human couldn't tell the respondent was an AI, then the AI was "intelligent", or at least intelligent enough. We've long since had programs which passed this test (back in 2014), but what you have not had is something general purpose and at scale.

ChatGPT and GPTs promise to bring this at scale.

If OpenAI is a software company that creates computer brains, and GPT-4 is the most powerful version of the brain, then the newly released "GPTs" are essentially Neo's cassette decks in the Matrix. They are configuration and context-setting for the brain, so that the computer brain is prepared for the conversation that's happening. In other words, GPTs help the machine "know what you mean" when you ask certain questions.

Why is conversation important?

Not everything can be solved by simple pages of text or videos. Sometimes, you wish you could have access to an expert on any topic or activity coaching or helping you through it, to clarify, or especially if the task is long-running and something goes wrong.

A conversation requires not just "stateless search" -- it requires context and state. Not just of the questioner, but also of the project. That is what LLMs can propose to solve.

The problem is, they are difficult to setup and configure. Language matters to the AI brain. That's what "prompts" are when you start using ChatGPT.

Enter GPTs: which anyone can create, use, and share with others.

These "GPTs" configuration files became a way to setup and use your own personal expert, and to the extent they are useful, other people can benefit from them. If they are very useful, you can sell them! Yes, GPTs are a new, perhaps generational, App Store opportunity.

GPT configuration is an expert on a particular set of activities or context that could be useful only within your own company (in a corporate setting), or globally (like a real WebMD).

In other words, GPTs can help you do anything, by helping you learn anything.

Your move, Google. As it stands, Google has been working to upgrade Bard to match OpenAI but has been somewhat one step behind.

It's difficult to overstate what's happening here, and helping you understand how to think about this helps you have conversations about this in your own company. Including if you are a retailer or brand.

Because the retailer that can teach consumers how or why to buy through conversation, has a head start on controlling what they buy.

Rick Watson

Rick Watson founded RMW Commerce Consulting after spending 20+ years as a technology entrepreneur and operator exclusively in the eCommerce industry with companies like ChannelAdvisor, BarnesandNoble.com, Merchantry, and Pitney Bowes.

Watson’s work today is centered on supporting investors and management teams incubating and growing direct-to-consumer businesses. Most recently, in partnership with WHP Global, Rick was a critical resource in architecting the WHP+ platform, a new turnkey direct to consumer digital e-commerce platform that powers AnneKlein.com and JosephAbboud.com.

Watson also hosts a weekly podcast, Watson Weekly, where he shares an unbiased, unfiltered expert take on the retail sector’s biggest players.

In the past year alone, Rick has spoken at many in-person and virtual events as well as podcasts on topics ranging from retail/ecom to supply chain/logistics and even digital grocery including CommerceNext IRL, ASCM Connect, and Retail Innovation Conference.

https://www.rmwcommerce.com/
Previous
Previous

5 Takeaways from Klaviyo's 3Q 2023 Results: Klaviyo-Sized Bullseye Painted on SMS Providers

Next
Next

Startups and Big Company Exec Before Product-Market Fit = Death