prior probability

Thinking like a machine (part 2 of 3)

Posted on August 17, 2016 by F. E. Guerra-Pujol

In our previous post, we mentioned John Danaher’s excellent review of Brett Frischmann’s 2014 paper exploring the possibility of a Reverse Turing Test. One of the insightful contributions Frischmann makes to this voluminous literature is his idea of a Turing Line, or the fuzzy line that separates humans from machines. According to Frischmann, this line serves two essential functions: (1) it differentiates humans from machines (and machines from humans, we would add), and (2) it demarcates a “finish line” or goal. In other words, for a machine to pass Turing’s original test, it must be able to cross this imaginary line by deceiving us that it is human. Most of the literature in this area focuses on the human side of the line: will a machine ever be capable of crossing this boundary? Frischmann, however, focuses on the machine side of the line. (In the words of Danaher: “Instead of thinking about the properties or attributes that are distinctively human, [Frischmann is] thinking about the properties and attributes that are distinctly machine-like.”) In particular, Frischmann poses a different and far more original question: will a human ever be able to deceive another person (or another machine) that he or she is a machine? But what does it mean to “think like a machine”? We shall discuss that difficult question in our next post …

Credit: Brett Frischmann (via John Danaher)

Posted in Bayesian Reasoning, Philosophy, Science | 4 Comments

Reverse Turing Tests and Ethical Machines (part 1 of 3)

Posted on August 16, 2016 by F. E. Guerra-Pujol

Our colleague John Danaher recently pondered the possibility of a “Reverse Turing Test” in this intriguing blog post dated 21 July 2016. That is, instead of testing for a machine’s ability to think like a human, what if we tested for a human’s ability to think like a machine? (This theoretical paper by law professor Brett Frischmann on “Human-Focused Turing Tests” is what led Danaher to pose this novel question.) Moreover, according to Danaher (and to Frischmann), the ability to think like a machine may have some serious ethical implications. For our part, we have often wondered whether ethical rules like Kant’s famous “Categorical Imperative” or the Golden Rule could be reduced to a simple computer program, and we have long been fascinated by the Turing Test; by way of example, we used the original version of Alan Turing’s famous test to develop the notion of “probabilistic verdicts” in this paper. Accordingly, we will be blogging about the ideas in Danaher’s post and in Frischmann’s paper over the next few days.

xkcd.com

Posted in Bayesian Reasoning, Philosophy, Science, Science Fiction | 1 Comment

Kant vs. Strauss vs. postmodernism

Posted on August 15, 2016 by F. E. Guerra-Pujol

If you had to choose, would you rather read 300 pages on Kantian nonsense, on Straussian esotericism, or on postmodernist garbage? Our colleague Jason Brennan, a philosophy professor at Georgetown University, wrote up this sarcastic taxonomy of the most common types of PhD dissertations in the fields of political philosophy and political theory, a comprehensive classification based on his personal experience of having served on many search committees for post-docs and junior candidates. (Hat tip: Brian Leiter.) The even-numbered abstracts were our personal favorites:

4. Incomprehensible Kantian Nonsense. “I’m going to argue that some policy P is justified on Kantian grounds. This argument will take 75 steps, and will read as if it’s been translated, or, rather, partially translated, from 19th century German. It will also be completely implausible, and so, to non-Kantians, will simply read like a reductio of Kant rather than a defense of P.”

6. Incomprehensible Postmodernist Garbage: “This dissertation examines the ontic-ontological ontology of late capitalist crises through the agonistic hyperrealist lens of soda dispensers and Fall Out Boy lyrics.”

8. Straussian Esotericism: “Here are three hundred pages written about the first two pages of Locke’s third letter to his second foot doctor. My dissertation does not defend any recognizable thesis, nor is it a piece of exegesis. Non-Straussians will have no clue what I’m doing. However, other Straussians will recognize it as deep.”

Posted in Academia, Bayesian Reasoning, Philosophy | 1 Comment

Bangladesh > Russia (population)

Posted on August 14, 2016 by F. E. Guerra-Pujol

Hat tip: Cliff Pickover

Posted in Bayesian Reasoning, Maps | 2 Comments

Additional critique of Baude and Sachs

Posted on August 12, 2016 by F. E. Guerra-Pujol

We mentioned previously that our colleagues Will Baude (University of Chicago) and Stephen Sachs (Duke University) posted to SSRN a fascinating paper titled “The Law of Interpretation” to be published in the Harvard Law Review early next year. In their paper, Baude and Sachs reframe the traditional rules of statutory and constitutional interpretation as a separate body of law. In the words of Professor Baude: “the law of interpretation can tell [judges] which of several contested theories to use when reading a text, what texts to interpret, what kind of background presumptions to use, and how to resolve uncertainties in those readings. Our model here is private law, where legal rules of interpretation are common and relatively uncontroversial …” In other words, instead of proposing a new method or micro-level theory of interpretation, Baude and Sachs claim that the existing rules of interpretation form a coherent body of binding law and that this “law of interpretation” is just as binding on judges as statutes and constitutions are. This is thus an ingenious macro-level theory, since the rules of legal interpretation already exist and since the analogy to private law (e.g., contracts, torts, etc.) will be familiar to all lawyers and judges.

Nevertheless, although we applaud Baude and Sachs for proposing a new way of looking at the problem of interpretation in law, there is an additional reason why we are skeptical about their macro-level theory. Simply put, even if there were a coherent and internally-consistent body of interpretative rules, and even if this body of rules were considered law, these rules are not really binding on judges in any meaningful sense. Why not? Because in addition to the potential problem of regress (which we discuss elsewhere), there is the problem of self-reference: judges are not only the ones who created the rules of interpretation; judges are also the same institutional actors who get to apply these rules and determine their meaning and application. Is there a way out of this circle?

Continue reading →

Posted in Law, Logical Fallacies | Leave a comment

#AcademicOlympics

Posted on August 11, 2016 by F. E. Guerra-Pujol

Credit: Crazy Scientist, PhD, via Twitter.

Posted in Academia | 1 Comment

Africa is larger than China

Posted on August 10, 2016 by F. E. Guerra-Pujol

Hat tip: Landon Schnabel, via Twitter.

Posted in Bayesian Reasoning, Maps, Questions Rarely Asked | Leave a comment

If the burroughs of NYC were separate States, they would get 18 electoral college votes

Posted on August 9, 2016 by F. E. Guerra-Pujol

NM & WV = 5 votes each; AK, MT, NS, SD, VT, & WY = 3 votes each (h/t: digg)

Posted in Bayesian Reasoning, Maps, Politics | Leave a comment

A silly recursive game

Posted on August 8, 2016 by F. E. Guerra-Pujol

Happy Monday! We stumbled upon the following pointless game via the Facebook page of our friend and colleague Dennis Wye Keen Khong:

It’s National Book Week: The rules: grab the nearest book to you, go to page 56, and copy the 5th sentence as your status. Do not mention the book. Post the rules as part of your status.

(This recursive game also appears in this Reddit thread.) Despite the utter pointlessness of this silly literary game, we find it completely contagious and irresistible! Here is our entry: