Peter Atkins, author of Atkins’ Physical Chemistry (11th edition, 2017) and Conjuring the Universe (2018), thinks that only science can tackle the big questions, and he suggests that AI will find the answers that stymie us:

The lubricant of the scientific method is optimism, optimism that given patience and effort, often collaborative effort, comprehension will come. It has in the past, and there is no reason to suppose that such optimism is misplaced now. Of course, foothills have given way to mountains, and rapid progress cannot be expected in the final push. Maybe effort will take us, at least temporarily, down blind alleys (string theory perhaps) but then the blindness of that alley might suddenly be opened and there is a surge of achievement. Perhaps whole revised paradigms of thought, such as those a century or so ago when relativity and quantum mechanics emerged, will take comprehension in currently unimaginable directions. Maybe we shall find that the cosmos is just mathematics rendered substantial. Maybe our comprehension of consciousness will have to be left to the artificial device that we thought was merely a machine for simulating it. Maybe, indeed, circularity again, only the artificial consciousness we shall have built will have the capacity to understand the emergence of something from nothing.

I consider that there is nothing that the scientific method cannot elucidate. Indeed, we should delight in the journey of the collective human mind in the enterprise we call science. Peter Atkins, “Why it’s only science that can answer all the big questions” at Aeon

Atkins is arguing that the fact that we do not understand what consciousness is, far from being a barrier to creating artificial consciousness, offers the hope that, once we do create them, artificially conscious entities will understand consciousness but we won’t.

The proposition sounds a bit confused, no? We will build a conscious machine even though we don’t know what consciousness is and it will understands consciousness even though we can’t?

The confusion is part of a more general problem. Naturalism (nature is all there is), often called “materialism,” is not working out well in science. Atkins mentions the frustrations of string theory but that is only one element of an entire cosmology where desire stands in for evidence.

The unpopular cosmos we are living in seems to have a beginning and to show evidence of design. The greatly desired cosmos, which came about purely by accident among an infinity of similar cosmos and shows no evidence of design, stubbornly refuses to appear in the data. The discussion is degenerating into accusations and name-calling.

This environment—not a pure abstraction of our own choosing—is the science environment we encounter. Let’s say we developed a powerful algorithm that we thought could understand the true nature of the universe where we don’t? How would we know it did?

Wouldn’t the algorithm beg likely to sound mysteriously like one of the quarreling factions? In that case, why should we believe that it represents the facts when we don’t really understand them ourselves?

If we must believe that “there is nothing that the scientific method cannot elucidate,” we could best preserve our faith by choosing only problems we can understand, to which the scientific method has obvious relevance. The current conflict over the nature of the universe is rooted in prior philosophical commitments and thus does not seem to be one of them.

See also: How naturalism rots science from the head down

and

Post-modern science: The illusion of consciousness sees through itself