Winter or Spring?

NLP is a sub-field of Artificial Intelligence. Many sub-fields of AI share with NLP a history in which periods of optimism give way to periods of pessimism. The leading term for this is "AI Winter". Expectations are set high, results are promised, funding comes from all corners, work begins, results fall short of goals, disappointment reigns, and funding goes away. Then, after enough time has passed, new expectations are formed and the cycle begins anew.

Of course, the cycles of failure are just part of the story. Speech-understanding has eventually found worthwhile (though limited) applications after surviving through cycles of hype and disappointment. Over time, the technology got better, the hardware got better and cheaper, and now it's rare to call a major corporation's customer support without being routed by a system that uses speech-understanding software.

But the cycles of failure keep coming, and they are continuing to the present day. Is NLP in Winter now or in Spring? Probably both. There has been a lot of funding for NLP-based ventures in the last few years. Some of it will lead to particular successes and some to particular failures. One beneficiary of high expectations of late has been San Francisco-based Powerset. Whether or not Powerset eventually delivers on the high expectations it has generated may have a lot to do with whether the field as a whole spends the next few years in spring or in winter.