Shayan Doroudi
1.3K posts

Shayan Doroudi
@EdTechMuser
Assistant Professor @UCIEducation Interested in [foundations/history/philosophy/equity] of ed tech, ed data science, and learning sciences

Since I was asked for an explanation: 1. It's not particularly surprising or impressive. 2. It's ugly, building basic objects from a baroque one. 3. It's useless. 4. It's foundationally and algorithmically pointless, since you still need to define exp and log, which you can't do without all the basic objects underneath. You can think of EML as a generator, but you still need the relations. As a 2-variable function, it is a syntactic building block for other stuff, but you still need to know which axioms it is subject to. So, yes, on paper it could be used to prove certain things by induction, but good luck finding some interesting property that is easier to prove on EML than on a relatively simple list of basic objects. And, no, it's not computationally useful, even less on silicon—good luck building a gate that implements that, and using it in a stable and packageable way.


God bless the departed soul of Professor Naquib Al-Attas. We are deeply in loss for his passing. His contribution to the Islamization of knowledge was unparalleled. It was Prof. Naquib's inspiration which laid the ground upon which I established the Islamia school in 1983. Prof. Syed Muhammad Naquib al-Attas was deeply involved in the 1977 First World Conference on Muslim Education in Makkah, the papers of which became the foundation for our curriculum research. May he join the great company of the prophets and learned as one of the shining exponents of God's Holy Book. #Peace








@SebastienBubeck @ChrSzegedy Personally I think trendlines are misleading in various ways; I’ve been saying for quite a while that existing benchmarks could be completely saturated without meaningfully touching many aspects of research math.
















