Since publishing 'Sales Ex Machina: How AI is Changing the World of Selling (2018), I've been getting this question:

"Victor, will AI replace salespeople?"  

It's the WRONG question!

Sales is not binary. It is a spectrum.

On one end sit pure transactions: low dollar value, low risk, high frequency, near-zero uncertainty. The buyer already knows what they want, the cost of a bad decision is trivial, and the entire process can be completed without a human ever speaking. 

On the opposite end sit high-stakes, high-complexity decisions: large capital outlays, multi-stakeholder politics, long implementation cycles, and real downside if the choice is wrong. Between those poles lives almost every sale that actually generates meaningful revenue.

When people ask whether AI will replace salespeople, the only honest answer is: it depends entirely on where the offer sits on that spectrum. I know, I sound like a lawyer. 😝 

Reality Reset: AI does not “replace sales.” It compresses the left side of the spectrum and raises the bar on the right.

The Spectrum Itself

Left Side is Simple / Transactional - Low risk. Low consideration. The buyer’s primary cost is time and convenience, not the price of being wrong. Trust is transferred to the brand, the interface, or the reviews (e.g., Amazon, Yelp, etc.). A salesperson adds friction, not value.

Right Side is Complex / Consultative - High risk. High consideration. The buyer is purchasing an outcome under uncertainty. They need diagnosis, risk mitigation, internal alignment, and someone who can still be held accountable after the contract is signed. Here, the salesperson isn't a talker; they are a risk-reduction mechanism.

Everything else exists as a gradient between those two poles. 

The further right you move, the more the sale requires human judgment, pattern recognition under incomplete information, and the ability to manage other humans’ fear.

Here are three examples of AI Sales On The Spectrum: 

1. Office supplies reorder (Far left) 

A facilities manager needs more pens, paper, and toner. The decision carries almost zero risk. The buyer already knows the SKUs, the preferred vendor, and the approximate price. AI-powered procurement systems, reorder algorithms, and marketplace interfaces already handle this better than any human ever could. A salesperson here is pure cost. 

2. Mid-market SaaS subscription (Middle of the spectrum) 

A 40-person marketing team is evaluating a $18k–$35k annual marketing automation platform. The risk is moderate: switching costs exist, data migration is annoying, and internal adoption is not guaranteed, but the company will not go bankrupt if the tool underperforms. AI can qualify the lead, generate a tailored demo environment, answer feature questions, and even produce ROI models. Yet, here's where GHOSTING comes in, the sale still frequently stalls on politics, competing priorities, and the quiet fear of “another tool that sits unused.” A skilled salesperson who can surface those unspoken risks and help the champion build the internal case still moves the needle. 

3. $500,000+ industrial automation cell or enterprise platform (Far Right) 

A manufacturer is deciding whether to invest half a million dollars (or more) in a robotic welding cell or a full ERP overhaul. The risk is existential: production downtime, integration failures with legacy systems, workforce resistance, and multi-year ROI that only materializes if the diagnosis was correct. No current AI system can walk the plant floor, notice the informal workarounds the operators have invented, read the political tension between the plant manager and corporate finance, or sit in the room when the CFO starts asking about residual value if the project fails in year three. The salesperson who functions as a true subject-matter expert and risk translator is still indispensable. AI can accelerate research, generate technical proposals, and model scenarios. It can't yet own the relationship or the accountability after the check is written.

AI does NOT flatten the sales profession. It polarizes it.

The left side of the spectrum is being automated at speed. The salesperson who survives is the one who can operate further right on the spectrum than the buyer’s own internal knowledge and the available AI tools can reach.

The spectrum does not care about job titles or headcount targets. It only cares about risk and uncertainty. AI is simply accelerating the separation between the two.

So, where do you or your company sit on the spectrum? The answer to that will answer the second-order question, "Will AI replace me?"

Victor Antonio, author of "Priced to Win"

P.s., Consider pre-ordering my new book, "Priced to Win: The Sales Leader's Framework for Closing Deals Without Caving on Price," due out in Nov. 2026 with Wiley Harper Publications.