Human Thoughts on ChatGPT - Part 2

Barry Bunin, PhD
Founder & CEO
Collaborative Drug Discovery

Can ChatGPT help businesses grow?

I see big ways that CDD (and thus other technology companies) can appropriately use generative AI technology such as ChatGPT to enhance their business pursuits.  In fact, I considered using ChatGPT for writing this article, but didn't want a machine to come between the reader and my own thoughts on the matter.  However, I separately added ChatGPTs thoughts on scaling our business and on CDD Vault's differentiation for comparison, as a P.S.

I think one of the real uses of the technology is for first drafts, especially when the writing doesn't require as much original human creativity.  In addition to creativity coming from the human mind, I'm interested in how ChatGTP can help in a sustainable way to augment human creativity in writing - what the founder of HubSpot, Brian Halligan, coined as "Remarkable Content" (content worth remarking upon in his first book on Inbound Marketing).

A table saw is a wonderful tool, but if you aren't careful you could lose some fingers using it. As companies begin to consider making use of ChatGPT, they should keep in mind the analogy from Ethan Mollick, a professor at the University of Pennsylvania's Wharton School of Business, who told NPR's Morning Edition:  "The best way to think about this is you are chatting with an omniscient, eager-to-please intern who sometimes lies to you." The story also noted: "It lies with confidence, too. Despite its authoritative tone, there have been instances in which ChatGPT won't tell you when it doesn't have the answer."

A colleague of mine said, "ChatGPT is great, but you have to remember that it can have the logical consistency of a befuddled kindergartener."  I think the key is in preparing ahead of time for how best to use the tool. Before interrogating ChatGPT, one needs to think of the right questions to ask, and afterwards, to critically edit and to thus validate the content. Perhaps most importantly, one can structure a logical framework or theme for presenting the results. In short, the results should be more than the technology can do alone, and ideally more than the human can do alone, too.

The technology will not replace people, but rather change the nature of the work requiring a human touch.

Derek Lowe, on his widely read In The Pipeline blog wrote, "I'm always talking about how we revise our definitions of such grunt work upwards over the years as machines become more capable, and this is the application of that principle to language.  Which makes me glad that I started blogging in 2002 and not now. I think (and this is a thought that many others have had as well) that LLMs [large language models] are going to end up devaluing a lot of human-produced writing, as it becomes apparent that it can be produced by machine, and simultaneously elevating whatever writing there is that can't be."

Using ChatGPT + PharmaKB to Create "Remarkable Content"

At CDD, we've created an application that will help produce remarkable content by leveraging our unique PharmaKB.com over time around the semantic content on the most recently FDA approved drugs-helping researchers and other drug discovery stakeholders to stay not just current but ahead of the curve on what's happening, and what could be happening, in this industry's extremely important work of bringing powerfully effective new drugs to the world.

AI's ability to ingest enormous data feeds and make high-value associations, means we can in a sense pool the current industry thinking-incorporating AI-generated insights and suggestions-and really help inform the world.  For example, consider the recent diabetes drugs, which are showing remarkable potential for obesity. From a recent Drug Discovery Roundup, I shared: "Novo Nordisk, the maker of the Wegovy injection that mimics the weight-related hormone GLP-1. This is a drug we've been following closely in PharmaKB 'The Pharmaceutical KnoweldgeBase' which we recently launched at www.pharmaKB.com. Because of the excitement over Lilly's drug Moujaro (tirzepatide) which also hits GLP-1 and a second target as mentioned in a recent FDA announcement I checked it out in PharmaKB, curious to see the top terms showing up in the literature for this new drug, based on the daily feed from PubMed Central."

With automation technology, and a little help from my friends, we can reproducibly create remarkable content around the launch of any new approved drug and the ramifications for all the competing drugs in the marketplace.

I see great potential to segment and customize messaging to be specific for each researcher. As with the scientific abstract (in part 1), remarkable targeted messaging content will need human intelligence upstream, and downstream to avoid being generic and to really resonate. If our general technology for handling Chemical and Biological Data evolves to support more use cases, someday we may even need to change our name from "Collaborative Drug Discovery" to "Collaborative Data Discovery".  As the CDD Vault technology co-evolves with broader swaths of the scientific community…we need to customize how we communicate the value for different applications:

https://www.collaborativedrug.com/cdd-vault-solutions-by-application/

Different users:

https://www.collaborativedrug.com/cdd-vault-solutions-by-role/

And adjacent markets:

https://www.collaborativedrug.com/cdd-vault-solutions-by-industry/

To paraphrase, Robert Anton Wilson, "those who are most intelligent, will use technology most intelligently."

Let the games begin.

Disclaimer: I personally have known and respected Sergey Brin since before Google went public. I especially respect his deep curiosity about technology I have bought and held Google stock for years.  I'm proud to say I have one of his earlier business cards from meeting at a Palo Alto Foresight Conference and subsequently at a Google SciFoo meeting.  I have a somewhat detached perspective as to how the stocks do and I am 99.9% focused on CDD Vault and been more recently playing with The Pharmaceutical KnowledgeBase.

P.S. For comparison, here is what ChatGPT wrote when queried with "How would you differentiate CDD Vault from competitors?

ChatGPT:  To differentiate CDD Vault from its competitors, it's important to highlight its unique features and benefits, and to show how it provides a better solution than other platforms in the market.

Here are a few key areas where CDD Vault can differentiate itself from its competitors:

1. Data Security: CDD Vault provides a highly secure platform for managing and sharing drug discovery data, with robust security features and controls in place to protect sensitive information.
2. Ease of Use: CDD Vault is designed to be user-friendly and intuitive, making it easy for researchers to access and use the data they need, without spending time on manual data entry or data management tasks.
3. Scalability: CDD Vault is designed to be highly scalable, allowing organizations to manage and share large amounts of data, and support multiple projects and teams, without sacrificing performance or security.
4. Collaboration: CDD Vault provides powerful tools for collaboration, enabling teams to work together more effectively, reducing duplication of effort, and accelerating the discovery and development of new drugs.
5. Customization: CDD Vault provides a high degree of customization, allowing organizations to configure the platform to meet their specific needs, and to integrate with other systems and tools.

Overall, CDD Vault can differentiate itself from its competitors by offering a combination of security, ease of use, scalability, collaboration, and customization, making it the ideal platform for managing and sharing drug discovery data.

=

Barry A. Bunin, PhD, is the Founder & CEO of Collaborative Drug Discovery, which provides a modern approach to drug discovery research informatics trusted globally by thousands of leading researchers. The CDD Vault is a hosted biological and chemical database that securely manages your private and external data.