Mettez-vous hors ligne avec l'application Player FM !
Vertical GPTs Will Change the Buyer's Journey
Manage episode 417132573 series 2359570
By now, I imagine most of you have interacted with ChatGPT in some way. You can use it to summarize a document, brainstorm a campaign or plan a vacation away from technology. The capabilities are mind-boggling and the BS (if you detect it) is amusing. All that aside, how can you use an internal vertical GPT to serve your customers?
I spoke to Ian Birkby, CEO of AZONetwork about Azthena, the internal vertical GPT they built as an AI assistant deployed on their various websites to help users find relevant news and product information. If you are not familiar with AZONetwork, Chat GPT describes them like this:
The AZoNetwork is a company that specializes in digital marketing and content solutions within the science, healthcare, and technology sectors. They provide a range of services aimed at connecting scientific, medical, and technology communities globally. These services include content creation, marketing strategies, and communication solutions designed to disseminate knowledge and promote products through various online platforms. AZoNetwork operates several websites that publish articles, news, and resources relevant to professionals in these fields, helping them to stay informed about the latest developments and technologies.
As a media and marketing company, AZONetwork has a lot of content, over a million assets, by Ian’s estimate. This is the content was used to train Azthena to to answer users queries.
The value of this type of GPT is that the data it has been trained on has been reviewed and vetted by humans in the course of publication over many years. The same can’t be said for everything one finds on the internet, some of which ends up in the answers to your ChatGPT queries, not to mention outright hallucinations.
General LLMs, if I'm going to exaggerate, are a mile wide and a few inches deep. We're trying to be, you know, maybe 12 inches wide, but three foot deep… - Ian Birkby
Azthena took nine months to build. That was impressive to me given that includes putting some structure around all of their content.
The quality of the output should enhance the customer experience by providing specific answers to queries along with relevant references.
If you find this helpful, it kind of makes sense to subscribe, doesn’t it?
Ian and I discussed what this will mean for companies, buyers and human creativity broadly.
External GPTs trained on the same broad collections of data may all end up homogenized. How will you differentiate your company to be a source of unique and valuable (essential) information? Building your own specialized GPT is one possible answer.
We agreed that the buyer’s journey will change. Once people adapt to the new approach (this will take some time but too much), they’ll expect to find answers and recommended next steps regarding what to look for or look at along the way to a purchase. No more clicking through multiple links hoping to find the best information. As I write that last sentence, I’m stunned at how quickly we are spoiled by new technologies. I’m old enough to remember mailing away to get a catalog or brochure from one vendor! If you are a marketer, you should be thinking about what comes next.
In reaction to this change, big information providers e.g. The Guardian, etc. are blocking crawlers from sucking up their information. Models built on search might also change. Will we have to pay for things that were free for the last 20 years?
Finally, I asked Ian about the future of human creativity.
…emotional intelligence… that's definitely an area where, you know, the human has still got a role I think, in that, very pure form of creativity… seeing what nobody else has seen before and thinking something different. You know, that I still think there's mileage in, in humans having a role to play there. However, there's probably 80 percent of the drudgery-related tasks that we all face that are going to end up sitting on the desk of AI.
But for those routine tasks, the efficiencies will be too attractive to pass up. The next question is where is the line that makes a task routine?
Your deepest insights are your best branding. I’d love to help you share them. Chat with me about custom content for your life science brand. Or visit my website.
This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit cclifescience.substack.com
209 episodes
Manage episode 417132573 series 2359570
By now, I imagine most of you have interacted with ChatGPT in some way. You can use it to summarize a document, brainstorm a campaign or plan a vacation away from technology. The capabilities are mind-boggling and the BS (if you detect it) is amusing. All that aside, how can you use an internal vertical GPT to serve your customers?
I spoke to Ian Birkby, CEO of AZONetwork about Azthena, the internal vertical GPT they built as an AI assistant deployed on their various websites to help users find relevant news and product information. If you are not familiar with AZONetwork, Chat GPT describes them like this:
The AZoNetwork is a company that specializes in digital marketing and content solutions within the science, healthcare, and technology sectors. They provide a range of services aimed at connecting scientific, medical, and technology communities globally. These services include content creation, marketing strategies, and communication solutions designed to disseminate knowledge and promote products through various online platforms. AZoNetwork operates several websites that publish articles, news, and resources relevant to professionals in these fields, helping them to stay informed about the latest developments and technologies.
As a media and marketing company, AZONetwork has a lot of content, over a million assets, by Ian’s estimate. This is the content was used to train Azthena to to answer users queries.
The value of this type of GPT is that the data it has been trained on has been reviewed and vetted by humans in the course of publication over many years. The same can’t be said for everything one finds on the internet, some of which ends up in the answers to your ChatGPT queries, not to mention outright hallucinations.
General LLMs, if I'm going to exaggerate, are a mile wide and a few inches deep. We're trying to be, you know, maybe 12 inches wide, but three foot deep… - Ian Birkby
Azthena took nine months to build. That was impressive to me given that includes putting some structure around all of their content.
The quality of the output should enhance the customer experience by providing specific answers to queries along with relevant references.
If you find this helpful, it kind of makes sense to subscribe, doesn’t it?
Ian and I discussed what this will mean for companies, buyers and human creativity broadly.
External GPTs trained on the same broad collections of data may all end up homogenized. How will you differentiate your company to be a source of unique and valuable (essential) information? Building your own specialized GPT is one possible answer.
We agreed that the buyer’s journey will change. Once people adapt to the new approach (this will take some time but too much), they’ll expect to find answers and recommended next steps regarding what to look for or look at along the way to a purchase. No more clicking through multiple links hoping to find the best information. As I write that last sentence, I’m stunned at how quickly we are spoiled by new technologies. I’m old enough to remember mailing away to get a catalog or brochure from one vendor! If you are a marketer, you should be thinking about what comes next.
In reaction to this change, big information providers e.g. The Guardian, etc. are blocking crawlers from sucking up their information. Models built on search might also change. Will we have to pay for things that were free for the last 20 years?
Finally, I asked Ian about the future of human creativity.
…emotional intelligence… that's definitely an area where, you know, the human has still got a role I think, in that, very pure form of creativity… seeing what nobody else has seen before and thinking something different. You know, that I still think there's mileage in, in humans having a role to play there. However, there's probably 80 percent of the drudgery-related tasks that we all face that are going to end up sitting on the desk of AI.
But for those routine tasks, the efficiencies will be too attractive to pass up. The next question is where is the line that makes a task routine?
Your deepest insights are your best branding. I’d love to help you share them. Chat with me about custom content for your life science brand. Or visit my website.
This is a public episode. If you would like to discuss this with other subscribers or get access to bonus episodes, visit cclifescience.substack.com
209 episodes
Tous les épisodes
×Bienvenue sur Lecteur FM!
Lecteur FM recherche sur Internet des podcasts de haute qualité que vous pourrez apprécier dès maintenant. C'est la meilleure application de podcast et fonctionne sur Android, iPhone et le Web. Inscrivez-vous pour synchroniser les abonnements sur tous les appareils.