Mustafa Suleyman is also DeepMind co-founder

In the article he seems to be saying several different things (maybe?)

1. We shouldn't give computers human emotions, or have them stimulate emotions. I guess because computers are better at solving problems and should leave human interaction to people.

2. We shouldn't try to make computers self aware. We should focus on making computers useful tools to people instead. Yet self awareness seems like it would be something that would make a program more useful...

3. Computers cannot be conscious because its just a simulation. I'd like to congratulate him on being the first and only person to understand consciousness, assuming it's even a well stated and understandable concept to begin with.

Yes, some things might be impossible, but there might be approximations that are good enough. I can't fly like a bird, but machines let me fly faster and higher than a bird. Maybe my computer can never be conscious, but perhaps it will eventually be able to do whatever I need from it as if it were conscious. Plus, history is full of naysayers later proven wrong, so please keep your "Can't" predictions to yourself.