|
A terrible puff piece. First of all, A New Kind of Science is mostly controversial (in scientific fields) not because what it says is controversial, it's not - actually sometimes trivially accepted as true - wait, what there are emergent phenomena that can be described by simple mathematics who'da thunk it? But rather because it almost attempts to redefine "science" which is somewhat dangerous. Search engines aren’t good at that, Wolfram argues, because they’re too messy. Questions in a search engine have many answers, with varying degrees of applicability and “rightness.” That’s not computable, not clean enough to program or feed into a system. “We want to be right,” Wolfram told me. “Making the world computable is a much higher bar than being able to generate Wikipedia-style information … a very different thing. What we’ve tried to do is insanely more ambitious.” This is exactly wrong, and perhaps Wolfram is a victim of what I like to call the "Lt. Cdr. Data fallacy" that sometimes afflicts AI - that AI is something which ought to our fuzzy world and assigns a definite binary result which is always or almost always going to be correct and perfect in every way. To take the example in the text, "Where is the ISS". There is not really one right answer. You could give precise siderially defined coordinates of the International Space Station, but, maybe "10 degrees to your left, and 20 degrees off your vertical axis in the downward direction, 12790km down" is a more correct and useful answer. Or, maybe if you're at the sandy spring transit station in Atlanta Georgia, "first right after the parkway on mount vernon" is correct; interpreting ISS to mean the former "internet security systems, inc" (which is a query that google "understands"). https://maps.google.com/maps?q=sandy+spring+transit+station+... Sometimes there just isn't one universally correct answer. Having said all of this, I have tremendous respect for Mathematica, which really helped me get through Grad school, mostly because their systems of differential equation solver is quite nice, and their graphing package was superior to excel. |