Categories
Articles

Top Twitter Topics by Data Scientists #24

Trending this week: Reinforcement learning and federated learning explained; AI is not equal to ML; Will AI become conscious; 9 ethical AI principles for organizations to follow.

Every week we analyze the most discussed topics on Twitter by Data Science & AI influencers.

The following topics, URLs, resources, and tweets have been automatically extracted using a topic modeling technique based on Sentence BERT, which we have enhanced to fit our use case.

Want to know more about the methodology used? Check out this article for more details, and find the codes in this Github repository!

Overview

This week, Data Science and AI influencers on Twitter have talked about:

  • ML Updates
  • Discussions On AI
  • Ethical AI

The following sections provide all the details for each topic.

ML Updates

This week, influencers have shared updates on machine learning and deep learning.

Carlos E. Perez has shared a link to a research paper titled “Revisiting the Primacy of English in Zero-shot Cross-lingual Transfer”. This paper explains that to build multilingual NLP systems, a successful recipe is to pre-train on a multilingual corpus, and then fine-tune on labeled data in a single transfer language — usually English. But, it also calls the choice of English for fine-tuning into question. Indeed, they compared English against other transfer languages for fine-tuning, on two pre-trained multilingual models (mBERT and mT5) and multiple classification and question answering tasks, and found that other high-resource languages such as German and Russian often transfer more effectively, especially when the set of target languages is diverse or unknown a priori.

Ipfconline has shared:

An article titled “Reinforcement Learning 101 — Experts Explain”. This post shares the thoughts of experts in the field on what the upcoming years could hold. The content of this post was gathered from members of the RE•WORK community including Lex Fridman of MIT, Thomas Simonini of UnityML Course, Doina Precup of UoM and more. This post provides the fundamentals of reinforcement learning (RL) and introduces deep reinforcement learning (DRL). Indeed, it starts by explaining the concepts of reward, may it be single of an infinite number, and then it also introduces the concept of Curiosity-Learning. Video gallery links are also provided on the footer page of this article.

A post providing an Introduction to Federated Learning (FL), a technique enabling on-device training, model personalization, and more. This post presents FL as a new type of training method for machine learning models that leverages ground-truth data generated by an end device (i.e. a mobile phone) to update a generic or shared model that’s distributed to different devices. This article explains: what FL is, what are the steps for FL, which properties of Problems can be solved using FL, and the federated averaging algorithm.

Discussions On AI

This week, AI and Data Science influencers have shared content talking about discussions on artificial intelligence.

Andreas Staub has shared a link to an article titled “AI Is Not equal to ML”. This post talks about the simplifications made by some people who keep talking about artificial intelligence (AI) while in fact, they are only referring to machine learning (ML). Firstly, it gives definitions of artificial intelligence and data science, which comprises machine learning, and then presents the different types of intentions in Business Analytics and the different technologies applicable for Prescriptive Analytics that help to understand the differences between AI and ML. Then, it concludes by saying that, for different types of AI intentions, there are different methods and techniques, that could use ML or not. This demonstrates well that AI and ML are not equivalent.

Marcus Borba has shared an article titled “Will Artificial Intelligence Ever Become Conscious?”. This post states that artificial consciousness is just a matter of time. It presents the hot topic of artificial intelligence developing consciousness as a pre-condition of sentient machines is the subject of speculative debates and fake AI start-ups profiteering. This post talks about the idea of putting a layer of emotional quotient (EQ) on top of AI, that is promoted as affective computing, with its commercial promoters. Then, it provides some statements on the capability of today’s AI/ML/DL systems to sense, feel, learn, know, which can open discussions.

On his side, Yves Mulkers has shared an article title “Artificial Intelligence vs. Human Intelligence: Which is the Force Majeure?”. This post starts by providing a definition of artificial and human intelligence. Then it compares them and states that even though while AI can perform repetitive tasks with greater efficiency, better accuracy, and higher speed, and enhance human intelligence in ways unseen before, human expertise is still the most crucial factor in designing AI technology in the first place.

Ethical AI

This week, Data Science and AI influencers shared their opinions and articles on ethical AI.

Bob E. Hayes shared an infographic on 9 ethical AI principles for organizations to follow.

The above infographic was also shared by Ronald van Loon. He also shared an article on OECD paving the way towards trustworthy and responsible AI. The article talks about worldwide dialogue and collaboration on AI and mapping AI policies and initiatives by country.

And finally, Tamara McCleary shared articles on five recommendations for creating more Ethical AI. The article talks about the importance of ethical AI and how it can be set up.

She also shared another article on Ethical Workplace & Artificial Intelligence. The article talks about the opportunities and risks of AI and how ethical AI can ameliorate the risks.