Neats and scruffies

In the <a href="/facts/History_of_artificial_intelligence/1wNLDHbB">history of artificial intelligence</a> (AI), neat and scruffy are two contrasting approaches to AI research. The distinction was made in the 1970s, and was a subject of discussion until the mid-1980s.
"Neats" use algorithms based on a single formal paradigm, such as <a href="/facts/Logic/8JnpJLtt">logic</a>, <a href="/facts/Mathematical_optimization/oRn8Iv5I">mathematical optimization</a>, or <a href="/facts/Neural_network_(machine_learning)/6V1jMlkx">neural networks</a>. Neats verify their programs are correct via rigorous mathematical theory. Neat researchers and analysts tend to express the hope that this single formal paradigm can be extended and improved in order to achieve <a href="/facts/Artificial_general_intelligence/vbxwaq8t">general intelligence</a> and <a href="/facts/Superintelligence/VL1NaEPS">superintelligence</a>.
"Scruffies" use any number of different algorithms and methods to achieve intelligent behavior, and rely on incremental testing to verify their programs. Scruffy programming requires large amounts of <a href="/facts/Hand_coding/zbvcTcuX">hand coding</a> and <a href="/facts/Knowledge_engineering/oQr2VuTu">knowledge engineering</a>. Scruffy experts have argued that general intelligence can only be implemented by solving a large number of essentially unrelated problems, and that there is no <a href="/facts/Silver_bullet/7aDhNYGU">silver bullet</a> that will allow programs to develop general intelligence autonomously. 
<a href="/facts/John_Brockman_(literary_agent)/VTDujWxU">John Brockman</a> compares the neat approach to <a href="/facts/Physics/a7aH2y08">physics</a>, in that it uses simple mathematical models as its foundation. The scruffy approach is more <a href="/facts/Biology/1M1ee3pB">biological</a>, in that much of the work involves studying and categorizing diverse phenomena.
Modern AI has elements of both scruffy and neat approaches. Scruffy AI researchers in the 1990s applied mathematical rigor to their programs, as neat experts did. They also express the hope that there is a single paradigm (a "master algorithm") that will cause general intelligence and superintelligence to emerge. But modern AI also resembles the scruffies: modern <a href="/facts/Machine_learning/e0w0XJTu">machine learning</a> applications require a great deal of hand-tuning and incremental testing; while the general algorithm is mathematically rigorous, accomplishing the specific goals of a particular application is not. Also, in the early 2000s, the field of <a href="/facts/Software_development/UUfsAaYY">software development</a> embraced <a href="/facts/Extreme_programming/UCeJIvhQ">extreme programming</a>, which is a modern version of the scruffy methodology: try things and test them, without wasting time looking for more elegant or general solutions.

Neats and scruffies open-in-new

Neats and scruffies