SoylentNews

We Need an FDA For Algorithms

posted by Fnord666 on Monday November 19 2018, @07:42AM   
from the who-watches-the-watchers dept.

upstart writes:

Submitted via IRC for Bytram

We Need an FDA For Algorithms

In the introduction to her new book, Hannah Fry points out something interesting about the phrase "Hello World." It's never been quite clear, she says, whether the phrase—which is frequently the entire output of a student's first computer program—is supposed to be attributed to the program, awakening for the first time, or to the programmer, announcing their triumphant first creation.

Perhaps for this reason, "Hello World" calls to mind a dialogue between human and machine, one which has never been more relevant than it is today. Her book, called Hello World, published in September, walks us through a rapidly computerizing world. Fry is both optimistic and excited—along with her Ph.D. students at the University of College, London, she has worked on many algorithms herself—and cautious. In conversation and in her book, she issues a call to arms: We need to make algorithms transparent, regulated, and forgiving of the flawed creatures that converse with them.

I reached her by telephone while she was on a book tour in New York City.

---

Original Submission

• Re:Can the Free Market solve this problem? (Score: 1) by khallow on Tuesday November 20 2018, @01:39PM

  by khallow (3766) on Tuesday November 20 2018, @01:39PM (#764234) Journal

  You are quite right – I’ve assumed there’s no regulation just cos I’m not aware of it, and she hasn’t mentioned any, but works in the field – so I’m just guessing she would have said if so... If there is regulation that you’re aware of, I’d be grateful if you’d link to some info?

  For example, there's plenty of cases where technologies or particular practitioners of technology have been ruled out as court evidence. So for the algorithm that's supposed to be used for sentencing? Show bias of the illegal sort (such as against gender, religious beliefs, etc in the US), and your have the basis for overturning every bit of sentencing done with that algorithm.
  
  Scratch the surface of almost any human activity and someone is regulating it. You just need to look.

  I don’t think we can pronounce her thesis as ‘stupid’ till we know more – not sure I can be arsed to read her book, but it is something that interests me, so maybe I’ll find the time! (I’m hoping there’s more detail than just ‘we need an FDA!’ - but also, highlighting a problem doesn’t necessarily require a solution to make it legitimate, surely?)

  It still needs to be a problem in the first place. The key flaw is simply that an algorithm is not an action. When you regulate algorithms, you aren't actually regulating the problem behavior. And as I noted earlier, there's already regulation (and means to implement more regulation should that become necessary) that don't require any sort of specialized regulatory system. People have been implementing bureaucratic algorithms, for example, for thousands of years.
  
  Further, most of the problems mentioned would be problems no matter what the algorithm was. Such as the project that siphoned a couple million UK residents' health data.

  I guess my fear is that the vested interests will object, of course – and they are powerful indeed. If the tech community doesn’t speak out about the dangers of relying on AI, then who will?

  What AI? It's not what we have now. And we don't know enough, in the absence of credible AI, to decide what aspects need regulation.

  Parent