Hi there! Welcome to the latest edition of The Sidekick! Today, we’re diving into the fascinating—and sometimes perplexing—world of generative AI.

From wild suggestions (non-toxic glue on pizza, anyone?) to its impact on procurement and supply chain, we’ll explore why a sprinkle of human insight is still crucial in this tech-driven age. 

That’s not all. Scroll down to check out more technology news, some curated highlights from social media, words of wisdom, tech news, people to follow, and more. 

Let’s dive in. 

If you’ve ever wondered how to get cheese to stick to pizza, don’t ask Generative AI. Google’s AI overview suggested that pizza enthusiasts should “add about 1/8 cup of non-toxic glue to the sauce to give it more tackiness.”

The same AI also confidently urged readers to “eat at least one small rock per day,” pointing erroneously to UC Berkeley geologists as its source. It also had the brilliant recommendation to “drink two quarts of urine every 24 hours” to help pass kidney stones.

Seriously…?

I feel like I remember laughing at a specific scene from Waterworld (1995); someone get Kevin Costner on the line - maybe he’s doing the coding.

These errors all point to the technology’s current limitations, which, depending on the use case, can be damaging and very dangerous. Take a business owner, for example, who decides to rely on Generative AI facts for a strategic decision.

What if someone has a medical emergency and asks ChatGPT what to do? Will the AI call an ambulance, or recommend the patient eats some pebbles?

The problem is that ChatGPT and its equivalents are language models, not knowledge models. People treat them like search engines, but that is definitely not what they were designed for. Take scientific research, for example, which is about observing the world and understanding its properties, then reviewing what we learn against what we already know.

ChatGPT can't do either of those things - it doesn’t observe reality or accurately represent existing knowledge. What it can do very well is mimic the style of a scientific paper, which can be handy. But it doesn’t take the place of actual scientific work.

GenAI makes mistakes when it doesn’t fully understand a term, or if a term sends it barking up on the wrong tree during a task. The result is a “hallucination,” which is why we need to be so cautious. It’s common knowledge at this point that OpenAI, Microsoft, Google, Apple, Huawei, and other players are scrambling to throw everything at AI without a clear plan - hence the chaotic, dangerous output. 

From Una’s perspective, we’re seeing an impact, too. We’re in the business of educating about group purchasing. We provide content that answers specific questions related to what we do and the services we offer. If people are searching for “group purchasing” and are given an AI response that has been mashed together (often with misleading or outdated information) that can directly impact their experience - and ours as a business. It becomes harder for people to find us, fact-check information, and make a decision about whether to join a GPO.

Recently, we noticed ChatGPT thinks Una (Sidenote, the models also like to capitalize our brand name to be UNA which is incorrect.) is a vertical Healthcare GPO, which isn’t true (we’re a horizontal GPO). Sure, we mention Healthcare GPOs in our brief history of group purchasing, but this is apparently enough to confuse the algorithm.

Press ChatGPT for answers about where exactly it is getting its information, and you’ll get an answer such as “I draw from a mixture of licensed data, publicly available information, and general industry knowledge.” 

In other words, GenAI is a black box. It doesn’t search for information like us humans. Its understanding is based on patterns in data, not a humanlike understanding of concepts that can be applied to new situations.

Let’s talk about using Generative AI in the world of procurement and supply chain. Relying on Generative AI for insights in these professions might not be the best idea for several reasons. First off, while AI can crunch a ton of data quickly, it doesn’t really understand the real world like humans do.

This means it can spit out recommendations that sound good on paper but miss the mark in practice. If the data it’s trained on is outdated or biased, the insights you get could lead you astray.

Another big issue is that procurement isn’t just about numbers; it’s also about relationships. Building trust and rapport with suppliers takes time and human interaction, something AI simply can’t replicate. If you rely too much on AI, you might miss out on those crucial conversations and connections that can lead to better deals or smoother negotiations.

Plus, there’s a risk of becoming overly dependent on AI. When organizations lean too heavily on data-driven insights, they might overlook important qualitative factors like supplier reliability, ethical practices, and changing market trends. This could lead to poor decisions that disrupt the supply chain.

If everyone starts relying on AI, it could stifle human creativity and expertise in procurement strategies. Adapting to a fast-changing market often requires intuition and experience - things AI just doesn’t have…yet!

We can’t ignore Generative AI. Every business needs to be on the AI curve or risk falling behind; however, there must be guardrails. It’s essential to balance it with human insight and judgement to truly navigate the complexities of procurement and supply chain management effectively.

What’s going on in the world of technology? Plenty. Here’s some news that caught my attention:

North Carolina man charged with AI-generated music fraud: In the first case of its kind, a man defrauded popular streaming platforms including Spotify and Apple Music by generating thousands of songs using AI, then programming thousands of fake accounts (again using AI) to stream the songs billions of times, ultimately earning $10 million in royalties. 

California’s agricultural sector could store water underground: A new study has recommended that instead of constantly building more dams and reservoirs (that are subject to evaporation), water should be directed into historically over-pumped and depleted aquifers below the Central Valley. 

Elon Musk’s Neuralink Blindsight device gets FDA approval: “Blindsight,” an experimental device that “will enable even those who have lost both eyes and their optic nerve to see,” has just won the FDA’s breakthrough device designation. Details are scant, but the approval may mean Neuralink is ready to move onto human trials.

If you’re looking to see how you can elevate yourself personally, I highly recommend checking out Chadd Wright. Chadd is a Navy SEAL veteran and the co-founder of “3 of 7 Project”. The 3 of 7 project has created training missions, runs a podcast, and has multiple resources to help you become the most complete version of yourself. 

What they are all about is encapsulated perfectly in their logo:

The "3" in 3 of 7 Project represents the body, soul, and spirit. The "7" represents a fully complete life where we lack nothing and have the tools and ability to help others and meet any challenge - physically, mentally and spiritually - head on.

3 of 7 makes up the skull of our logo, representing how the body, soul, and spirit make up our complete person. The skull also represents the bones of our mission - to bring life and purpose to those who have gone dry by giving real life applications on strengthening every area of our being.

The bird on top of the skull represents new life and new ideas coming from the mind and is positioned for takeoff.

"The forest was shrinking but the trees kept voting for the axe; for the axe was clever and convinced the trees that because his handle was made of wood, he was one of them." 

- Turkish Proverb

Please keep the suggestions coming; the responses and feedback have been great. If you’d like to see more, please hit the “subscribe” button below and please share with a friend. 

See you next time!

– Kris

Connect with me on LinkedIn!