Want more time, money, and energy from AI? Do not ask how many words it can make. Ask how many people open, trust, and act on those words.
That is the quiet lesson inside this week's loud AI news. On July 9, OpenAI released GPT-5.6. Its top model finished a broad set of intelligence tasks in 61% less time than a rival frontier model, at about half the estimated cost. The machine can now think, browse, code, and build at a speed no team can match.
Then a new workplace study landed on July 13. It found that AI-written emails did not raise open rates or reply rates on their own. What did? The emotional tone inside the message.
The model got faster. The human signal still got the reply.
The machine made output cheap
For most of business history, making the answer took time. A report took days. A deck took a week. A good email took thought. Companies paid for people who could find facts and turn them into words.
That price is falling fast. GPT-5.6 can produce stronger knowledge work for less money, and its lower-cost models bring that power to more people. Information is becoming close to free. So is the first draft.
This sounds like pure profit. It is not. When every worker can make ten messages, every customer and teammate must sort through ten times the noise. Output rises, but attention does not. Speed moves the bottleneck from creation to connection.
This is where psychology meets the balance sheet. A message that feels cold gets skipped. A plan that feels forced gets slow-walked. A smart answer from a leader people do not trust becomes one more tab to close. The cost shows up as lost hours, weak follow-through, missed sales, and tired teams.
The reply comes from the human layer
The July 13 study followed 121 workers at six companies for three weeks. They sent 16,880 real work emails. Some were written with no AI. Some were rewritten by GPT-5 in a playful tone. Some were rewritten in a professional tone.
Neither AI condition directly changed opens, replies, or response time. But more positive language within the same sender strongly predicted action. It was tied to about twice the odds of an open and 3.32 times the odds of a reply. That is not a promise that cheerful words will triple every result. It is a strong signal that how a message feels can matter more than how it was made.
Higher emotional positivity was linked to 3.32 times the odds of a reply in a July 13 field experiment covering 16,880 workplace emails. AI use alone had no direct effect on replies.
Now connect that to trust. SHRM surveyed 5,875 U.S. workers in March and April 2026. Among workers who trusted their company and leaders to act with integrity and stay true to the mission, 67% had higher job engagement. Among workers without that trust, only 6% did.
The same rule runs through email, culture, and profit. People act when they feel a human signal they can trust. Tone opens the door. Character keeps it open. Emotional intelligence reads the moment. Clear purpose tells people why the action matters.
Measure Return on Individual
The old ROI asks what the tool produced. The new ROI asks what the individual turned that power into.
Did AI save an hour, or did it fill the hour with more noise? Did the faster proposal win the deal? Did the manager use the saved time to coach a person, solve a conflict, or go home with energy left for family? Did the message protect peace and partnership, or just move another task?
This is Josh's trillion-dollar EQ thesis in plain words. AI made information free. Information is not the moat anymore. Actual Intelligence, the human operating system, is the moat that is left. Character aims the tool. Critical thinking checks the answer. EQ makes it land. Grit carries it through. Elastic intelligence connects the right facts to the right human need.
Companies do not get paid when AI makes words. They get paid when a person says yes, makes a choice, solves a problem, or trusts the team enough to tell the truth. The best operator will use machine speed to create more profit, passion, peace, partnership, and purpose, not more volume.
Bottom line: the scarce asset is no longer the answer. It is the human who can turn the answer into action. Measure that human layer before you buy another layer of software. Start with the free Actual Intelligence diagnostic, then run the free Culture Test at actualintelligenceos.com.