By The Chronicle Staff

Artificial intelligence is already changing how people write emails, answer customer questions and communicate at work. But what we're seeing today may be only the beginning.

The bigger transformation could come when AI stops being something people intentionally open and begins operating quietly between nearly every conversation.

Imagine receiving a three-paragraph email and never actually reading it. Your AI assistant summarizes it in two sentences, recognizes that the sender needs an answer by Friday and places the issue on your task list.

You tell the assistant what you want to say, and it prepares a response in your normal writing style.

The person on the other end may have a similar assistant. Suddenly, two people could exchange several emails without either person actually writing — or perhaps even reading — most of them.

That sounds futuristic, but pieces of that future already exist. AI-powered email systems can summarize long threads, draft responses and help prioritize messages.

AI meeting assistants can record discussions, create summaries, identify action items and generate follow-up tasks. The next step may be moving from AI-assisted communication to AI-to-AI communication.

Consider something as ordinary as scheduling a meeting.

Today, one person might email, "Does Tuesday afternoon work?" The other checks a calendar and responds that 2 p.m. doesn't work, but 3:30 p.m. does. Someone sends an invitation. Another person asks to move it. Several messages later, everyone finally agrees.

In the future, much of that conversation may disappear. You might simply tell your AI assistant, "Schedule an hour with John sometime next week."

Your assistant could communicate with John's assistant, compare calendars, understand each person's preferences, select a time, reserve the meeting and provide both people with the information they need beforehand.

Humans wouldn't necessarily communicate less. They might spend less time on communication that exists primarily to coordinate other communication.

Email itself could become something completely different. Instead of opening an inbox containing 87 unread messages, you might begin the morning by asking, "What do I need to know?" Your AI could explain that a customer needs an answer, an employee is waiting for approval, three newsletters contain nothing important to you and 15 routine messages have already been handled.

Instead of organizing communication around individual messages, AI could organize it around decisions.

Meetings are another obvious target. AI assistants can already transcribe conversations, produce summaries and identify action items.

Eventually, attending every meeting personally may not be necessary. An AI assistant could potentially attend certain meetings on your behalf, understand the discussion, answer basic questions based on information you've authorized it to access and return with a summary of anything requiring your decision. 


A two-hour meeting might become a five-minute briefing.

That could also change who needs to attend meetings. Instead of 15 people sitting through an hour-long discussion because each person needs one small piece of information, perhaps five people participate while everyone else's AI assistants collect what their owners need.

One of AI's most promising effects on communication may be reducing language barriers. Real-time translation is improving rapidly.

A business owner in Ohio could eventually speak English during a video call with a manufacturer in Japan while the manufacturer hears Japanese. The manufacturer could respond in Japanese while the American hears English. Neither person would necessarily need to speak the other's language.

Similar technology could become common during phone calls and even face-to-face conversations through earbuds or other devices.

That could dramatically expand who businesses can hire, sell to and work with. A small company in a rural community could potentially communicate with customers and suppliers anywhere in the world almost as easily as it communicates with someone across town.

AI could also make mass communication increasingly personal. Today, a business might send the same email to 5,000 customers. Tomorrow, AI could potentially create 5,000 variations.

One customer might receive a short explanation because AI knows that person prefers concise messages. Another could receive more detail. Someone who speaks Spanish could receive the information in Spanish. A customer who prefers video could receive a video explanation instead.

The underlying information remains the same, but its presentation changes for the individual receiving it.

All of this efficiency creates an important question: If AI writes our emails, responds to customers, summarizes meetings and communicates with other AI systems, how do we know when we're actually communicating with another person?

That distinction could become increasingly valuable.

A handwritten note may mean more because we know someone actually wrote it. A telephone conversation may carry more weight because we know another human is on the other end. An in-person meeting could become more meaningful precisely because it can't be automated.

AI-generated communication also creates questions about authenticity. When anyone can produce a professional email, realistic voice recording, photograph or video in seconds, simply seeing or hearing something may no longer be enough to establish that it's real.

People may increasingly want verification. Was this message actually approved by the person whose name appears on it? Is that really the CEO speaking in the video? Did a journalist actually conduct this interview? Is the customer-service representative a person or an AI?

Trust could become one of the most valuable currencies of the AI era.

Perhaps the most interesting possibility is that AI doesn't eliminate human communication. Instead, it eliminates much of the communication humans never particularly wanted to do.

Scheduling meetings, summarizing conversations, rewriting awkward emails, taking

notes, translating routine information and answering repetitive questions consume enormous amounts of time.

If AI handles more of that work, people could theoretically spend more time on conversations requiring empathy, judgment, persuasion, creativity and relationships.

The technology becomes responsible for moving information efficiently, while humans remain responsible for determining what that information means and what should happen next.

That distinction may ultimately define communication in the AI era.

The telephone allowed people to communicate across distance. Email allowed messages to travel almost instantly. Smartphones made communication nearly constant. Artificial intelligence could represent the next major change.

But instead of simply giving us another way to communicate, AI may become an intelligent layer sitting between nearly every method we already use.

The future may not be humans talking to computers. It may be computers helping humans understand one another — while increasingly talking to other computers so we don't have to. ■