Is Generative AI Just Glitter Or Is It Gold? Could We Be the Lab Rats?

With every invention, there is some human collateral damage as the invention improves. These people are called first movers, and are essentially and unfortunately, the lab rats.

There is a lot of buzz around ChatGPT and generative AI, but many are not taking a second to step back from the excitement.

This may sound contradictory to my chosen career path as a Venture Capitalist in AI and entrepreneurship, but it's really not. My job as a leader is to ensure that I am promoting AI for good, and if I cannot do that, then my job is to deconstruct bad AI from what it is, and reconstruct it for the good it could be. Since 2007 I have been working with AI. Today I am an AI Venture Capitalist, Serial Entrepreneur, and I also run an accelerator focused first on AI. As I said, my journey with AI started in 2007 when I ran a command operations center for the Marines, where I reported to agencies like the Pentagon. Furthermore today as an AI Venture Capitalist and Board Advisor to Ithaca College for their cyber security program, I have uncovered multiple needs for governance and leadership around AI. Since the majority of the market is creating one AI product after another, this has, of course, resulted in an explosion of security challenges, and economic and social factors needing to be addressed and governed by the collective leadership of advisory boards, councils, and other factions. We do not need more AI products right now as a globe, we need to take a step back and look at how we are governing and leading the trends around the growth of AI and the data it uses. This is why part of the reason for this article and it is why I would like to continue to contribute and get involved

We must realize that generative AI in its current state is nascent and emerging, and there are still a lot of gaps to be figured out and addressed yet. Who wants to use their most precious information and help the big Silicon Valley companies and the big brother government figure it out? Not me, especially as far as the government is concerned, I already did my service to them in the U.S. Marine Corps!!

This is another level of power you're giving up than simply sharing data on Facebook or other social media. So judgment and caution is recommended when deciding what information to share with ChatGPT and other generative AI's, especially if you do not want your organization’s confidential information used and repurposed by the technologies (which goes back to my thoughts that we need more governance and leadership within AI, not more products)

Generative AI systems like ChatGPT, do generate content but is it as reliable as the hype seems to make it?

Even when references are annotated or cited the technologies can be heavily biased and assumptive.

While the intent of ChatGPT is noble and needed, the technology is potentially exacerbating existing problems, not solving them. Problems like inaccurate information, falsities, enabling hackers, phishing attacks, etc.

Prime example I gave a presentation recently that was titled “Mitigating Limitations Around AI”. During this presentation, I did a competitor test on ChatGPT with YouChat AI (the supposedly best ChatGPT alternative), and I asked both platforms “What Will the Population of Chicago Be In 2030.”

Each platform gave me starkly different answers yet based on the same data from the U.S. Census Bureau See the figure below (snapshot and results below from my presentation).

CHATGPT

The population of Chicago by 2030 answers (based on US census bureau data) - 10.5mm - 11mm people.

YOUCHAT AI

The population of Chicago by 2030 answers (based on US census bureau data) - 9mm people.

So, you can see the drastic contrast between the 2 data sets I collected. There is a contrast in Chicago’s 2030 population of 1.5mm and 2mm. Not only that, the YouChat AI claims the population in 2030 of Chicago will be 9mm, but ChatGPT claims it is already at 9.5.mm as of 2021. So which is it? Is Chicago losing or gaining here by 2030? What the hell is going on?

It is obvious we have an issue. On the ChatGPT result, we can clearly see from one of the best AI's on the market, the limitation and lack of accuracy. In addition, when it comes to more serious matters, this may result in dire consequences for economic, social, and governance efforts due to real deficits in bias and inaccuracies.

If you do decide to use generative AI platforms, I have provided a list in the link below of generative AI platforms for each being best at various use cases and scenarios. Scenario based usage and choosing the best platform for the situation may also help mitigate vulnerabilities with generative AI. In other words you wouldn't tow a trailer with a Ferrari as that is not what Ferrari cars are used for by consumers. Make your pick but don’t be a lab rat, use it judiciously!!

List of the top Generative AI platforms

This is what you should be thinking about when using Generative AI

Thank you for reading my thoughts, and feel free to learn more about me via my CV site - https://paulclaxton.io

#chatgtp #youchat #generativeai #artificialntelligence #aichatbots #aibias #ai #aigovernance #leadership #venturecapital #openai #openaichatgpt #microsoft #labrat #firstmover