ESP Biography



AARON KALB, Geek / Teacher / Language-Lover




Major: Symbolic Systems

College/Employer: Stanford

Year of Graduation: 2011

Picture of Aaron Kalb

Brief Biographical Sketch:

"English In, Answers Out." That's the vision of the company I co-founded—with a couple Google engineers and an Oracle VP—in 2012, after working on Siri at Apple. I've been fascinated by how people and machines process languages throughout my five years and two degrees at Stanford, where I studied Symbolic Systems (Computer Science + Linguistics + Logic + Psychology + Philosophy). I'm excited to share some of what I've learned with a new batch of Splash students this Fall!

In my spare time I love hiking/camping, playing board games, traveling, eating Thai food, and making/solving puzzles.



Past Classes

  (Clicking a class title will bring you to the course's section of the corresponding course catalog)

M4056: Google Search and Google Translate: How Humans (Accidentally) Teach Computers in Splash Fall 2014 (Nov. 08 - 09, 2014)
Suppose you have billions of books, and someone asks you for all the pages containing the word “eggplant”, sorted by the quality of the books according to the world’s eggplant experts. How long would it take you to deliver the results? If you answered longer than 300 milliseconds, then you have something to learn from Google. It takes a human longer than that to even read—let alone to answer—such a question. In that time, Google actually fulfills thousands of such requests. Their process is brilliant: conceptually simple but technically complex, and in this fast paced tour, we’ll zoom through the basics of both, covering: - PageRank - Inverted Indices - TFIDF - Text alignment - Language modeling and more… We’ll end by discussing some limitations of the Google approach to Search and Translation and hinting at next steps in both fields…


R3331: Making Meaning: Sounds, Symbols, and Semantics in Splash! Fall 2013 (Nov. 02 - 03, 2013)
If you're reading this, your brain (with a little help from your eyes) is engaged in an elaborate process of turning black squiggles on a white background into conceptual understanding. Language--written, spoken, and signed--has completely transformed the way humans function, within society and within our own minds, and is largely responsible for our triumphs as a species. In this fast-paced, interactive, and interdisciplinary course we'll explore questions like: - How and why do languages evolve and change? - What does it mean for two words or phrases to "mean the same thing"? How do humans and computers attempt translation between languages or paraphrases within a language? - Are there some concepts that can only be expressed in certain languages? How do the languages we speak shape the way we think? We'll also decode ancient manuscripts, learn how whole religious movements stemmed from mistranslations, and translate Dr. Seuss.


M3332: Google Search and Google Translate: How Humans (Accidentally) Teach Computers in Splash! Fall 2013 (Nov. 02 - 03, 2013)
Suppose you have billions of books, and someone asks you for all the pages containing the word "eggplant", sorted by the quality of the books according to the world's eggplant experts. How long would it take you to deliver the results? If you answered longer than 300 milliseconds, then you have something to learn from Google. It takes a human longer than that to even read, let alone answer, such a question. In that time, Google actually fulfills thousands of such requests. Their process is brilliant: conceptually simple but technically complex, and in this fast paced tour, we'll zoom through the basics of both, covering: - PageRank - Inverted Indices - TFIDF - Text alignment - Language modeling and more... We'll end by discussing some limitations of the Google approach to Search and Translation and hinting at next steps in both fields...


S829: Games in Splash! Spring 2010 (Apr. 17 - 18, 2010)
What is a game? How can we come up with a meaningful definition which encompasses board games, computer games, war games, mind games, and love games? Why are games fun? What makes games hard? If a computer can play a game well, is it intelligent? Can a computer enjoy playing a game? In this interdisciplinary course (placed here because I have to put it somewhere) we'll explore these questions and more, drawing from research in Game Theory (economics), Game Design, Artificial Intelligence, Modal Logic, and Linguistics, among other topics. We'll also spend some time making and playing games.


S404: Why? in Splash! Spring 2009 (Apr. 04 - 05, 2009)
Why? Why what? Why learn logic? Because it's awesome! And that argument is valid, though it might be unsound. Why? You'll find out when you take this course, over the course of which (no pun intended) you will: - Embark on a whirlwind tour of Formal Logic and its myriad applications - Dance (metaphorically) at the intersection of the humanities and sciences, exploring issues that form of the basis of philosophy, computer science, linguistics, mathematics, and (perhaps) all human thought. (This last point depends on findings from psychology, which we'll also cover). - Read and discuss op-eds and political speeches, identify their underlying (il)logical structures, and then rip them to shreds Stretch your brain by solving cool puzzles and tackling deep questions Have engaging, possibly heated debates - Learn all about: > The reasons for reason: the issues that arise with natural language and untutored human problem solving, and how formal logic can help remedy them. > Bullshit: how to spot it and refute it, how to use it and get away with it, and--best of all--how to argue without needing it > Propositional, Modal, and First-Order Logics, their notations, histories, strengths, weaknesses, and uses. > The Limits of Logic: why if you are obligated to be gentle rather than cruel, then you are obligated to kill your mother (and why that may or may not be a problem) > What this means: $$ \forall x \Box ShouldTakeThisClass x $$ ...and why it's true. By the end of the course you will: - Become a more discerning reader, consumer, and thinker, immune to the ploys of demagogues and advertisers, - Be able to win any argument... (with a computer at least, humans tend to rely on emotions--something else we'll discuss), - Be 3.7 times smarter than you were before, - Know whether this course description is mere rhetorical flourish or a sound and valid argument for taking this course, and most importantly - Be able to answer that most profound of all human queries and most detested of all follow-up questions: Why?