If it is true that...

AI could narrow our range of inquiry

and

It would be evil to narrow our range of inquiry

Then it must be true that...

AI has the potential for evil outcomes

If it is true that...

AI has the potential to deepen isolation

Then it must be true that...

AI has the potential for evil outcomes

If it is true that...

If we don't actively build AI for good, its evil outcomes may manifest

and

AI has the potential for evil outcomes

Then it must be true that...

We must actively build AI for good