It has always seemed to be that a hackathon is ideal for producing two kinds of things:
- Trivial things that perhaps might have market value — kind of like the “pet rock” or the “Chia Pet” — but that society really doesn’t need; and/or,
- Things that an organization (the one sponsoring the hackathon) should have produced through thoughtful means, but couldn’t because it is so internally stifled, that the only way to liberate the innovative ideas of its staff is through a chaotic hackathon.
Was Google created in a hackathon? No: the algorithms were developed by two researchers at Stanford. The algorithms were primary; the code was secondary.
And a-lot of really impactful things take more than a hack day to create. SpaceX’s first rocket blew up. So did the next. And the third. The fourth worked, and years later they proved that you can reuse boosters.
I am personally not interested in what “coders” can cobble up in a day or two of rushed activity. It is not likely to change the arc of society.
For example, there has been a-lot of progress in computer algorithms in the past few decades, but those algorithms came from research — not from hacking. Machine learning is a case in point: Hinton’s algorithms were derived mathematically before they were coded, and the deep CNN networks that have proven so effective for image recognition were not the result of hackathons: they were created by PhD researchers who focused on their models and then coded them as step 2.
Hackathons are not how real innovation happens.
