One of the interesting quotes floating around in this week’s news cycle is from Mark Cuban. In the segment, he says “the value of a computer science degree will ‘diminish over time’”. The full quote is:
“Twenty years from now, if you are a coder, you might be out of a job.. Because it’s just math and so, whatever we’re defining the AI to do, someone’s got to know the topic. If you’re doing an AI to emulate Shakespeare, somebody better know Shakespeare … the coding major who graduates this year probably has better short-term opportunity than the liberal arts major that’s a Shakespeare expert, but long term, it’s like people who learned COBOL or Fortran and thought that was the future and they were going to be covered forever.Mark Cuban, SALT 2019. Interviews linked here.
There are a lot of things which could be said about this, but I’d say just two:
- Cuban is right the software engineering itself is no more “secure” against future change from automation than other disciplines. However, it’s also true that, both technically and in terms of deployment in the world, there is a fantastically long way to go before this happens. AI systems can already rewrite parts of their own code (you could do this in Lisp and Smalltalk in the 90s) but it’s incredibly hard (and frankly not profitable) to do it outside of learning algorithms today. There are still vast numbers of manual and physical things which need to be augmented with technology before we start to see a decline in the need for software engineering skills.
- What is more important about this quote is the emphasis on not giving up our “real world” skills: Shakespeare expertise (his example), medicine, legal skills, investment judgments, teaching, etc. It is more and more the case that success in the next 20-30 years will depend on a combination of real world domain expertise and the use of improving technical tools.
In other words, it is in bringing simple yet powerful tools to practitioners in their fields of expertise and those practitioners getting good at using them that we’ll get the most positive (and equitable) outcomes from technology.
One example of automation in a field does not spell the end of a field. The question to ask when this happens in your area is: how can technology be simplified and improved to augment and make more valuable the existing humans who know their stuff?