Why Retailers Shouldn't Worry About the OpenAI Senate Hearings
It’s clear that artificial intelligence is here to stay. Soon, every piece of software and industry will have AI in it. But, naturally, it’s scary. There's so much unknown and the recent Senate hearings on regulating AI have brought up new fears across every industry. Will jobs get replaced? Why would I ever trust this to make critical decisions? And perhaps most importantly, will my data be safe?
Businesses, including high-profile companies like Apple, have even banned staff from using conversational AI tools like ChatGPT. These same fears are present in the retail industry. However, blanket bans are short-sighted and will eventually be detrimental to any business. The power of conversational AI will continue to push forward. In fact, by 2033, global AI in the retail sector is expected to achieve a $85 billion valuation. There’s no going back. Instead of facing AI with fear, retailers need to move forward and can do so safely.
You Can Embrace AI for Your Supply Chain, Safely
Conversational AI models, if trained and applied effectively, can empower human decision makers to have a "conversation" with technology so they can make the best possible decisions — without sacrificing nuance. There are many ways that this can be valuable to a retail business. For example, imagine if the individuals who have to decide where a product gets disposed of could type in a query and get the best possible answer. Or if the individuals managing your inventory could simply ask what products are running low and be fed back an answer that's tied specifically to your business and based on your internal business data. All within seconds.
At present, decision making surrounding examples like these is left to one individual or a small team. These people are tasked with understanding a vast amount of information. With the right mindset and resources, you can be using conversational AI to improve their workflows — confidently and safely.
Remember, it’s No Different Than Storing Your Data in the Cloud
Bottom line: if you're comfortable storing your data in a database hosted on a cloud platform, then you shouldn't worry about using these AI models. Large language model (LLM) AI vendors have similar data storage and usage controls to that of database hosting. It would be no more work for Google to take data out of your cloud-hosted database than for OpenAI/MSFT to use your chatGPT sessions. It’s not something to worry too much about, but it’s important to remember that the data privacy concerns with AI are no different than those you're likely already familiar with.
Familiarize Yourself With All LLM AI Models
There are options out there, and each really is different. While ChatGPT is surely the most well-known, other LLMs exist. These include Google Bard, Microsoft Azure OpenAI, and various open-source LLMs. The number of options will surely grow, offering even more solutions. It’s important to explore the nuances of each, choose one that works best for your business, and structure how you implement it internally in a way that's safe and makes sense for your team.
Make Sure You Understand and Implement Usage Policies
That being said, you should absolutely start by first understanding AI model data usage policies and how they differ across each. Most options have policies that offer in-depth information about their use of data and, more often than not, are enterprise friendly. However, different options have different controls and policies. MSFT Azure OpenAI service, for example, is by far the more enterprise-friendly implementation of GPT-based APIs. And OpenAI has another service called "Foundry" that provides a completely private instance of GPT. Depending on what model is used in your business and by your employees, you'll need to create internal company policies and processes to make sure your team is very clear on how to use each — safely.
Still Not Sure? You Can Spin Up Your Own
At the end of the day, these AI models are open source. So if all of these steps and reasons still don’t have you feeling comfortable, you absolutely can spin up your own LLM. It's absolutely feasible to develop your own LLM and train it. And this is the safest way you can take advantage of the benefits of conversational AI without having to worry about data privacy.
Ultimately, conversational AI can and will be essential to your business. It won't replace your workforce (digital products don’t save companies; people save companies), but it will be a powerful tool for making your team’s lives easier, your decisions more accurate, and your processes more efficient. Retailers unwilling to adapt will be left behind in the wave of technological growth.
Charlie Vallely is co-founder and chief innovation officer at SmarterX, a regulatory services partner to retailers, brands and manufacturers. The company helps to accurately, and quickly, classify products for safe disposal and shipping.
Charlie Vallely is co-founder and chief innovation officer at SmarterX. SmarterX is a regulatory services partner to retailers, brands, and manufacturers. The company helps to accurately, and quickly, classify products for safe disposal and shipping.
Charlie champions novel ways of solving problems with a focus on the application of ML and AI to make the world a better place. He’s a passionate, hands-on leader who works closely with his team and clients, dedicating his days to efficiently scaling massive datasets and connecting data sources in the service of better product classifications.
His mastery over data science, computational chemistry, and hazardous waste regulation is almost entirely self taught - and might be the only thing he cares about more than his fantasy baseball team, which is saying a lot. Prior to SmarterX, Charlie created, launched and managed an analytics software company and ran a digital consultancy firm.