Talk is becoming the new user interface

Salesforce recently published an excellent article exploring how AI is changing the way we interact with software.

One quote in particular stood out:

“When he typed a prompt, he tidied his thoughts before hitting enter. When he spoke, however, he rambled, stammered, doubled back — yet the results improved. Freed from the instinct to clean things up, Cherny gave the machine more of his actual thinking, which gave the AI more to work with.”

It’s a simple observation, but I think it points to a much bigger shift in enterprise software.

We’ve been adapting to computers for decades

For years, we’ve translated what we know into formats that software can understand.

We fill in CRM fields. We write meeting notes. We summarise conversations. We complete forms. We deliberately simplify rich, nuanced conversations into structured records because that’s how enterprise software has always worked.

The software needed structure, so we provided it.

But AI changes that relationship.

Conversation contains more than data

Think about a typical customer meeting. Someone explains why they’re hesitant to move forward. A competitor is mentioned. Budget concerns emerge halfway through. The customer changes their mind, circles back to an earlier point, then finally reveals the real objection.

Very little of that richness survives when someone later writes: “Customer raised pricing concerns.”

It’s precisely that nuance that helps humans understand situations, and increasingly, it’s also what AI needs.

Voice is becoming a better way to capture context

Speaking is different from typing. When we talk, we naturally elaborate, clarify, reconsider, and connect ideas as we go. That often gives AI far more context than we’d ever include in a manually written update.

The challenge, of course, is that conversations are inherently unstructured. Enterprise systems like Salesforce rely on structured data. Historically, that gap has required people to manually interpret conversations and translate them into CRM updates.

That’s where AI becomes genuinely transformative.

From voice to structured business data

The interesting innovation isn’t speech-to-text. It’s speech-to-business-data.

Imagine leaving a customer meeting and recording a 30-second voice memo while walking back to your car.

As well as producing a transcript, AI identifies the relevant Salesforce records, updates opportunity fields, creates follow-up tasks, drafts an email, and captures the context alongside the customer record.

No typing. No forms. No “I’ll update Salesforce later.” Just natural conversation becoming structured business intelligence.

Why this matters for Agentforce

This isn’t simply about saving time. It’s about improving the quality of information available across your organisation.

AI agents don’t just need records. They need context.

The richer the conversational history attached to an Account, Opportunity, Case, or Contact, the better AI can understand what’s happening, identify risks, recommend actions, and automate meaningful work.

For many organisations, conversations represent the largest source of untapped customer intelligence. The question is no longer whether AI can capture them.

It’s whether that intelligence becomes part of your system of record.

A different way of thinking about CRM

For years we’ve thought about CRM as something people update after the conversation. Perhaps that’s about to change. Perhaps conversation itself becomes the update. Perhaps speaking becomes the interface.

It’s an exciting shift because it removes one of the biggest bottlenecks in CRM adoption: asking busy people to stop what they’re doing and become data entry clerks.

The organisations that embrace this change won’t simply have less admin.

They’ll have richer customer context, better organisational memory, and a stronger foundation for the next generation of AI.

The Salesforce article that inspired this post is well worth reading. It explores how AI is changing the way we interact with technology and why conversation is becoming an increasingly important interface for enterprise software.