"AI Can't See!" The AI Phase After Agentic That Will Reshape How Dealers Run Every Department (+ How to Use It) | AJ McGowan, VP of Research and Development at Reynolds & Reynolds

Reynolds and Reynolds is a major dealership software company used across thousands of dealerships in the U.S. The episode frames it as a key platform where AI features can be deployed across sales, service, and other departments.

AJ McGowan is the guest in this segment and is positioned as a technology leader tied to Reynolds and Reynolds’ AI research direction. The discussion uses his perspective to explain the evolution from generative to agentic to cognitive AI for dealership operations.

“ChatGPT” (spoken here as “chat GBT”) is a conversational generative AI system built on large language models (LLMs). It’s relevant to dealers because it can help automate and speed up customer communication and internal drafting, but it needs guardrails to avoid unreliable outputs.

Sam Altman is the founder/co-founder figure associated with OpenAI, mentioned here in connection with the viral early reception of ChatGPT. The hosts use his example to illustrate how AI products can face skepticism before becoming widely impactful.

OpenAI is the organization associated with ChatGPT, referenced here as the company behind the AI system that later “revolutionized” many things. In this segment, OpenAI is used as context for how generative AI entered the mainstream.

Cognitive AI is described here as the next phase after generative and agentic AI—aiming to be truly transformative. The idea is that it should better understand context and make more reliable decisions, rather than only generating text or executing simple automated tasks.

Agentic AI refers to AI systems that can take actions toward a goal, not just generate text. In dealership software, “agents” can trigger tasks automatically inside tools like a CRM or DMS—such as updating records, following up with leads, or performing workflow steps.

DMS stands for Dealer Management System. It’s the dealership’s core software for managing operations like inventory, service/repair workflows, and customer records—so AI agents can automate operational tasks there.

An omnichannel buying experience means customers can shop and buy across multiple channels (like website, phone, email, and in-store) with a consistent process and information. The host frames this as what dealers want software to support, not a single “next Carvana” app.

Carvana is an online used-car retailer known for a heavily digital sales funnel and a “buy online, deliver to you” approach. The host uses it as a shorthand example of a single-channel, app-driven model that dealers aren’t necessarily trying to copy.

Machine learning is a type of AI where systems learn patterns from data to make predictions or decisions. In the transcript, it’s framed as an earlier phase of AI compared with newer approaches like generative AI and LLMs.

“Cognitive software” is a broad label for software designed to mimic aspects of human thinking—like understanding language, interpreting information, and supporting decisions. In the transcript, it’s used to describe the evolving phases of AI that dealers might leverage across departments.

ChatGPT is a widely used conversational AI system built on large language models (LLMs). The host uses it as a reference point for how the transformer/LLM approach brought generative AI to the mainstream.

LLM stands for large language model—an AI trained on huge amounts of text to predict and generate language. In the transcript, the LLM is presented as the core capability behind generative AI tools that can summarize, draft, and respond.

A transformer is a neural-network architecture that became the foundation for many modern language models. It helps models understand relationships between words in context, which is why it’s closely tied to LLM performance.

Generative AI is AI that creates new content—like text, summaries, or drafts—based on a prompt. The host contrasts it with earlier “bounded task” uses, where the AI’s job was more limited and predictable.

The AMC Matador is a mid-size car made by American Motors Corporation (AMC) that was produced mainly in the 1970s. It’s often discussed in automotive history because it represents how American automakers built larger, more comfort-focused cars during that era. In a podcast context, it may come up as a recognizable name tied to classic-car culture or dealership storytelling.

Auto Vision is described as a data-science product that helps dealers value cars more accurately. The host frames it as providing data and analysis that support pricing decisions, then adds an inventory platform around it to make it easier for dealers to use.

An inventory management platform is software that helps a dealership track and manage its vehicle stock—what’s on the lot, what it’s worth, and how pricing decisions affect turnover. In this context, it’s built to make the company’s car-valuation data usable inside the dealer’s workflow.

A Bloomberg Terminal is a well-known financial data and analytics system used by professionals to access market information and make decisions. The host uses it as a metaphor: Auto Vision is positioned as a similar “data hub” for used-car dealers, with historical trends and valuation comparisons.

Avery is the name of an AI agent the speaker says was built to use Auto Vision’s tools for car valuation. The described goal is to recommend a specific pricing number and related actions, based on the same evaluation process a user would follow in Auto Vision.

Ray is described as a “genetic platform” made of a series of AI agents inside an architecture. The key idea is that it can use multiple products in the Reynolds ecosystem and perform a search/optimization across data sources to answer dealer questions with an objective target.

Goal seek is an optimization/search approach where the system adjusts inputs to reach a target outcome. In this transcript, Ray is said to “go and goal seek” across data sources to answer users’ questions and achieve an objective number.

CRM is customer relationship management software. In a dealership context, CRM data is used to track leads and customer interactions—like how many leads come in, how they’re handled, and what converts into sales.

To “reprice” a vehicle means adjusting its listed price, typically in response to market movement, demand, or inventory aging. The segment frames it as an automated action that could happen daily based on AI-driven recommendations.

VEC in this context refers to a vehicle evaluation/pricing capability used to estimate what a car should be worth. The segment says dealers gain “insight into what their VEC is doing,” implying it influences behavior like pricing and appraisal decisions.

Appraisal time is when a dealership evaluates a vehicle’s condition and value—often for trade-ins or used-car pricing. The segment notes AI can run recommendations at this stage, which can speed up inspections and improve pricing decisions.

The service drive is the area/process where vehicles enter for maintenance and repairs. The segment uses it as an example of proactive AI: predicting incoming service jobs and identifying missing parts before the cars arrive.

Inventory acquisition is the process of sourcing vehicles for resale—through buying cars, trades, auctions, or other channels. The segment suggests large companies already automate parts of this process, and that automation will spread over time.

“Systems talk” refers to integrating dealership software so different departments’ tools can share data and coordinate actions. The goal is that AI can use information from multiple systems (service, customer messaging, sales/CRM) to make a single, coherent recommendation.

“Contextual databases” are data systems designed to store and retrieve information with meaning tied to a specific situation (like a particular car, customer, and time). In a dealership setting, that lets AI combine service history, lease status, and customer preferences to generate relevant next steps instead of generic prompts.

A “lease” is a financing arrangement where the customer pays to use a vehicle for a set term, rather than owning it outright. The segment highlights lease timing (“two payments left”) as a key piece of data AI can use to identify when a customer is likely to consider a trade or purchase.

In dealership context, “warranty” usually refers to coverage sold alongside the vehicle—often extended coverage beyond the factory term. The speaker’s example treats warranty add-ons as part of the AI’s recommended offer when a customer is likely to be shopping soon.

A “trade” in dealership language typically means trading in the customer’s current vehicle toward the purchase of another vehicle. The segment frames trade readiness as an AI insight derived from lease status, customer behavior, and vehicle relevance.

A “service rep” is the dealership’s service department representative (often the service advisor) who communicates with customers about maintenance and repairs. The example emphasizes that the AI should understand customer communication preferences and service interactions to coordinate a handoff to sales.

This is a principle about AI systems: they can only act on signals they can access (e.g., data feeds, integrations, and user-permissioned inputs). In dealership operations, it implies that if customer/vehicle context isn’t captured in connected systems, AI can’t “notice” it and can’t personalize outreach.

In dealer software, CDP usually refers to a customer data platform—tools that collect, organize, and unify customer and lead information. The point here is that the AI needs access to the right customer/contact data to act on opportunities across departments.

“Agentic” refers to AI systems that don’t just answer questions—they take actions on your behalf (for example, triggering outreach, coordinating tasks, or moving a lead through a workflow). In this context, the “egenic/agentic future” framing emphasizes that AI will need real-time access to dealership data across departments to act effectively.

Opportunity cost is the value you give up by not choosing the best alternative. The speaker uses it to argue that keeping dealership software/data fragmented will become increasingly expensive over time because AI won’t be able to see everything it needs, reducing automation and responsiveness.

In dealership tech, “consolidated” means centralizing software and data so multiple departments share the same information rather than working from disconnected systems. The segment ties this to AI effectiveness: if data isn’t consolidated, the AI can’t connect leads, service activity, and sales context in real time.