Winter isn’t coming. At least not for the world of bogus intelligence. Jeff Bigham, affairs administrator of the Human-computer Interaction Institute at Carnegie Mellon, believes we’ll soon reap the bounties of an AI autumn instead.

An AI winter is a period in which analysis and allotment freeze up and it becomes difficult to break new ground. Since the birth of bogus intelligence technology in the 1940s there’s been several of these periods. The last decade has brought new hope for AI – and with it all the hyperbole the modern media can muster. AI is in the public’s eye on a larger scale than ever before, this much is certain.

So it would then follow that greater hype would lead to a colder winter. Bigham seems to accept this won’t be the case, because we’re absolutely accomplishing stuff with modern AI. In a recent blog post he says:

AI hype is abbreviating all around us, and what will be left is a rich autumn of human-centered abstruse work applying Apparatus Acquirements to important problems.

If AI autumn is a period where we’ll start seeing some return on the hype cycle, then what absolutely does that mean? Today’s neural networks are already saving lives, bearing art, and making billions of dollars for big tech companies. Are we about to enter a golden-era of angelic AI super human symbiosis?

Nah, not really. Bigham says it’s all about human-computer interaction:

People who can apply ML to solve real human problems will become the most important tech people out there. Powerful ML is more captured in easy-to-use libraries; if you want to stay ahead of the curve, you need the skills we teach in our HCI curriculum.

If your goal is to fight through the winter, with hopes of anytime developing truly able AI, then break free from the stranglehold of deep acquirements and applied application, and go forth bravely.

If your goal is to reap the rewards of the harvest, study HCI.

He goes on to point out that it’s become more common for developers to accede the human factor too late into the assembly of a apparatus acquirements system. His altercation for human-centric AI analysis predicts that useful AI will begin to rise to the top as consumers, the media, and the public at-large eventually lose absorption in over-hyped pipe dreams from people and companies able things beyond the adeptness of today’s accepted apparatus acquirements technologies.

According to Bigham, the next phase of modern AI – its autumn — will allow accomplished developers to prosper. To quote him just the aboriginal bit out of context, it’s become accessible this phase will be about apprehension the problems that are “both important enough to matter” and “narrow enough” that specific machine acquirements methods will work well. Unfettered by out-of-whack expectations during this modest-to-accurately-hyped season, advisers will assuredly have time to abode the field’s abiding open questions.

Unless he’s wrong about the whole winter thing. Then we’ll all watch as the AI bubble bursts and bags of startups and major technology companies go under amid broker panic.


Read next: Notre Dame fire sparks Twitter cabal theories