Story perspectives

Microsoft Pushes “Humanist Superintelligence” With Guardrails

By Drooid · Sep 20, 2026, 2:09 AM UTC · How we work

Summary

Story summary

Microsoft AI chief Mustafa Suleyman says the company is moving toward “humanist superintelligence,” serving people within limits.
OpenAI announced an automated “research intern” system, aiming for a fully automated researcher by March 2028.
Elon Musk said xAI’s Grok models will have humans “less in the loop” on improvement.
He added full automation could arrive by end 2026, not later than 2027.

History

Historical Context

In 1997 IBM’s Deep Blue, a 30-processor RS/6000 SP that evaluated 200 million chess positions per second, beat world champion Garry Kasparov 3½–2½, earning a $700,000 prize and proving brute-force AI could surpass human expertise. That milestone set a precedent for today’s efforts by Microsoft’s Mustafa Suleyman, OpenAI and Elon Musk’s xAI to build fully autonomous systems, reviving debates over superintelligence governance.

1 of 2

Story summary

Microsoft AI chief Mustafa Suleyman says the company is moving toward “humanist superintelligence,” serving people within limits.
OpenAI announced an automated “research intern” system, aiming for a fully automated researcher by March 2028.
Elon Musk said xAI’s Grok models will have humans “less in the loop” on improvement.
He added full automation could arrive by end 2026, not later than 2027.

1 / 2