Did you see ChatGPT coming? I sure didn’t and most of the AI experts I know were caught off-guard, too. One day, we all suddenly had access to one of the world’s most knowledgeable intellects, and we’re still scrambling to figure out how to get the most value from large language models (LLMs).
This quest will continue for a long time as models become more powerful and we learn how to interact with this exciting and yet still-mysterious intelligence.
In our company, COO Brian Hopkins has been the leader in building LLM tools for our team, including overlaying Claude onto our company data in a set of “Projects.” This allows us to complete work faster and better than ever.
He’s also created “prompt generators” because as it turns out, ChatGPT, Perplexity, Gemini and other tools are better at writing prompts for themselves than we are creating them on our own! I tell the LLM what I’m trying to do, and it asks me a series of questions. It uses the answers as a prompt to give me better, more useful results than I would have gotten if I’d just typed my own question or instructions.
The key limiting factor in LLMs is that they’re inherently passive. When I ask them questions or assign a task to them, they usually do an amazing job for me. Write a proposal? Claude looks at our previous proposals, considers my detailed request (which it helped to write with a prompt generator), evaluates our company’s capabilities and writes a proposal – complete with pricing – all in a few minutes.
However, LLMs don’t currently work across applications and datasets and proactively do work for me. They respond to what I ask them to do, always ready, but never interacting with other technologies or jumping in on their own before I ask them to.
Up Next: Agentic AI
Brace yourself because the next big thing in AI is coming soon and it’s called, “Agentic AI.” The name comes from AI acting as an agent on your behalf. Here’s a simple example:
I need to fly to Dallas for a presentation to the American Supply Association on June 18. With Agentic AI, as I’m driving back from the store, I can trigger the voice AI on my phone and say to the agent (which I’ll refer to as AIN – my name, spelling updated to the world of AI):
“Hey AIN, schedule my June ASA trip.” The agent will book my flights, rent me a car and confirm my hotel reservation. It will know the airlines I prefer, make sure it registers everything with my loyalty numbers, remember that I don’t like my return flight to be less than three hours after I finish speaking and add all the information to my calendar.
It can generate travel directions, monitor for weather delays and adapt automatically; it can find a gas station for me to fill up the car before I return it. It will even check the dress code and technology requirements for the event and generate a packing list for me. As I travel, it will automatically track my receipts and send in the expense statement afterwards.
AIN will have access to my travel accounts, Office 365 account, credit cards and work databases. It will know or learn my preferences, watch for any changes that come in by email or voicemail and, in exchange for all this authority, make my travel life easier and better than ever.
That’s a very simple example. Now think of all the ways you could use this kind of capability in your business, such as:
- answer customer questions
- manage your supply chain
- negotiate contracts
- watch for customer service issues before you or the customers know they’ve occurred
Imagine asking your agent to prepare your company as a snowstorm approaches. It will check inventory levels on winter products and report back to you minutes later with specific products and locations where you need to act – or act itself to optimize inventory allocations and order more stock based on manufacturer availability. It will send notifications to your team to watch for travel delays and generate a marketing email to your customers to tell them how you can help them get through the storm.
The use cases are endless because AI agents will have access to enormous amounts of data, will be able to consider many alternatives, balance many dependencies and – most importantly – act on your behalf. You won’t have to schedule calls; if someone asks for a meeting, you’ll simply tell your agent to set it up and it takes care of it for you. The change from reactive to proactive means agentic AI’s impact on your personal productivity will be even larger than LLMs.
Challenges in Agentic AI
Like LLMs, Agentic AI will pose challenges due to its immense intelligence and power. LLMs still hallucinate and sometimes provide wrong answers; Agents will make wrong choices and mistakes, especially when they’re early in the learning curve.
There are ethical questions to answer. If an agent replies for you, is it really you? Are its agreements legally binding? As these systems become more powerful, it’s harder for us to control them – are we willing to cede that kind of authority to technology?
Tough questions, to be sure, but we humans tend to adopt technologies that are convenient despite the risks – just look at the early days of commercial air travel; we were flying on jetliners before there was a radar grid and air traffic controllers!
And keep in mind the differentiating characteristic of AI vs. all previous technologies: It learns. It improves on its own. It gets better at its tasks without human intervention. Whatever shortcomings agentic AI comes with, it will improve over time; the trajectory will be sharply upward as it accumulates more data and experience.
The Bottom Line …
… will be the bottom line. As long as agents simplify our lives, make us more productive, lower our stress levels and make us more money, we’re going to use them. The advantages will be enormous, and you need to learn about agentic AI now to accrue those benefits sooner than your competitors.
Thinkers360 just published their Top 50 Global Thought Leaders and Influencers on Agentic AI 2025. At the top of the list, with a score of 100 (2nd place was a 66) sits Noelle Russell from the AI Leadership Institute. Noelle is joining us at this year’s Applied AI for Distributors in Chicago on June 24–26. She’s delivering a keynote presentation and will join us for the reception afterwards so you can ask her questions and learn from her. Noelle has vast expertise on AI in general and Agentic AI in particular and we’re honored and delighted to welcome her to the conference.
We’d like to welcome you, too! If you want to keep up, catch up or stay ahead in AI, you or someone from your company should attend this conference. Many companies send multiple attendees so they can attend simultaneous sessions on different tracks. This is the only major AI conference for distributors in the U.S. – don’t miss it.
Learn more about Applied AI for Distributors.
Ian Heller is the Founder and Chief Strategist for Distribution Strategy Group. He has more than 30 years of experience executing marketing and e-business strategy in the wholesale distribution industry, starting as a truck unloader at a Grainger branch while in college. He’s since held executive roles at GE Capital, Corporate Express, Newark Electronics and HD Supply. Ian has written and spoken extensively on the impact of digital disruption on distributors, and would love to start that conversation with you, your team or group. Reach out today at iheller@distributionstrategy.com.