Around the Web #9

DALL·E 2, divine simplicity, and Ronaldo’s ordeal.

1: [Movies] 🎥🍿🎬

I: Totems

I saw the series Totems on Amazon prime video. I liked it.

“In 1965, in the midst of the Cold War, scientist Francis Mareuil prepares to work for the French secret services and the CIA; he meets Lyudmila Goloubeva, a pianist forced to work for the KGB, and falls in love with her.”

II: Aníkúlápó

Set in the old Oyo empire of West Africa,  Aníkúlápó is a 2022 Nigerian epic fantasy film produced by Kunle Afolayan

“After an affair with the king's wife leads to his demise, a zealous traveler encounters a mystical bird with the power to give him another chance.”

Bifarin V  @BifarinTheFifth

Saw #Anikulapo on @NetflixNaija yesterday, apart from the obvious argument against crass indiscipline. Another one is that if you benefit from the brazen indecency of a friend, just note that when the situation presents itself, the same brazen indecency will be unleashed on you.

Netflix Nigeria @NetflixNaija

A match made in heaven or a pair doomed to fail? Aníkúlápó is now streaming 🌪️ https://t.co/EJY2URBPQa

2. [Biomedicine] 🩺🥼🧑🏻‍⚕️

I: I recently completed a paper on interpretable machine learning and metabolic phenotyping. In English: I showed that in the study cohort investigated, I can distinguish between males and females using machine learning by leveraging metabolites in their urine. And more importantly, I explained using machine learning methods what is driving such differences. Preprint.

Olatomiwa Bifarin, Ph.D. @BifarinPhD

🚨Pre-print alert🚨 I am glad to share my paper on interpretable machine learning in #metabolomics. I applied tree-based SHapley Additive exPlanations (SHAP) for a more ML interpretability depth. A 🧵 (1/7) [biorxiv.org/content/10.110…]

biorxiv.orgInterpretable machine learning with tree-based shapley additive explanations: application to metabolomics datasets for binary classificationMachine learning (ML) models are used in clinical metabolomics studies most notably for biomarker discoveries, to identify metabolites that discriminate between a case and control group. To improve understanding of the underlying biomedical problem and to bolster confidence in these discoveries, mod…

II: An explanation for age-dependent COVID-19 severity. Paper

III: Fungi ecologies in cancer. Paper

“Cancer-microbe associations have been explored for centuries, but cancer-associated fungi have rarely been examined. Here, we comprehensively characterize the cancer mycobiome within 17,401 patient tissue, blood, and plasma samples across 35 cancer types in four independent cohorts. We report fungal DNA and cells at low abundances across many major human cancers, with differences in community compositions that differ among cancer types, even when accounting for technical background.”

IV: Applying machine learning to biology is hard. I can attest to this. Endless confounders to worry about, instrument variations, big-p little-n problems, limited knowledge about biology itself, et cetera.

3: [Web3/Blockchain] 🌐🌐🌐

I: The web3 creator playbook: an essay.

"Instead of creating content for free with the hopes of gradually growing an audience and one day being able to monetize it, creators are now able to monetize and build an audience upfront through tokens"

4: [AI] 🤖🖥👨🏿‍💻

I: A neat substack on categorizing interpretable Machine Learning methods by Christoph Molnar. I will also recommend his book too. it taught me most of all I know about IML.

II: Make a video via text courtesy MetaAI. text-to-video generation without text-video data

See Paper, See the project page.

III: On GPT-3, zero-shot learning, perils, and promises. Beginner-friendly, short essay

IV: DALL·E 2 is an AI algorithm that generates images just from natural language (text). I played with it recently.

V. Sam Altman X Greylock Video

Promises of large language models, AI alignment problems, Multimodal models, continuous learning models, metaverse and AI, tools for creatives, etc.    

VI. A good machine learning book. 

Bifarin V  @BifarinTheFifth

I have read bits and pieces of this book. It's really good.

Frank Nielsen @FrnkNlsn

"Mathematics for machine learning" (CUP, 2020). Free PDF from https://t.co/mbzGgyFDXP 😊 A must-read masterpiece, nicely structured and richly color-illustrated! https://t.co/2V14nY7L2n

VII. Large language models crossed with biology and chemistry from NDVIA? Just beautiful! I worked on a project in grad school that aimed to predict the toxicity of a compound on a model organism with its molecular structure. However, the AI system I built did poorly. This LLM NDVIA is about to release might make things work from my end. See article

5: [Nigeria] 🇳🇬

I: An excellent take on Sowore’s bid for the Nigerian presidency so far, and his foray into politics in general.

“Sowore should be scrutinised and judged fairly, a tradition with which he’s of course familiar. But laughing at his bid is finding our misery funny.”

II: See my latest bite-sized book review on the making of Nigeria, and I can’t recommend the book enough for students of Nigeria.  

III: If you are of the Yoruba extraction or if you understand Yoruba, This is really good narration about Nigeria before Nigeria: Ìtàn Nàìjíría on Alaroye via YouTube.

IV: Donald Duke's brilliant analysis of PDP, Nigeria. On Arise TV.

6: [Philosophy/Theology] 🧠🧐✝️

I: Great content on divine simplicity, creation, etc., by Fr. Thomas White.

The lecture was given on July 17, 2022, at the Dominican House of Studies in Washington, D.C., for the Fourth Annual Thomistic Philosophy and Natural Science Symposium: Complexity, Simplicity, and Emergence.

Prerequisite: Thomistic metaphysics.

7: [Sports] 🏀 ⚽️ 🏈 

A nice essay on Ronaldo and his ongoing ordeal at Manchester United. Cristiano Ronaldo and the long walk.

“Soccer’s biggest stars used to have places to wind down their careers with dignity. The sport’s economics now require their humiliations be public.”