Categories
Articles

Top Twitter Topics by Data Scientists #13

Trending this week: Release of GPT-Neo 1.3B by EuletherAI; Train deep learning models on advanced mathematical calculation examples; Scale models to billions of parameters in a few code lines with Pytorch Lightning.

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 Sentence BERT based topic modeling technique that we have enhanced to fit our use case.

Want to know more on the methodology used? Jump into 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:

  • Deep Learning Updates
  • Impact of AI In Human Life
  • Covid-19 Vaccination Program In The US

Here are all the details for each topic:

Deep Learning Updates

This week, the influencers have provided links to resources on deep learning updates.

Hugging Face has published about the release of two checkpoints for GPT Neo, last week, in their Transformers library. This model named GPT-Neo 1.3B is a transformer model designed using EleutherAI’s replication of the GPT-3 architecture.

Still talking about EleutherAIYves Mulkers also tweeted about an article title “This AI Can Generate Convincing Text — and Anyone Can Use It” that explains how the EleutherAI’s project aims to “be an open source alternative to GPT-3.

Photograph: CASPAR BENSON/GETTY IMAGES

On his side, Yann LeCun shared the link to the the source code, datasets and trained models on Github related to a research paper by Facebook AI titled “Learning Advanced Mathematical Computations from Examples”. This publication demonstates how “neural networks can learn to perform complex computations, grounded in advanced theory, from examples, without built-in mathematical knowledge”, just by “using Transformers over large generated datasets”.

Finally, Andrej Karpathy shared an article showing how to use Pytorch Lightning, a lightweight PyTorch wrapper for high-performance AI research, to make it accessible and very simple to train deep learning models with multi-billion parameters. “This post introduces the PyTorch Lightning and DeepSpeed integration demonstrating how to scale models to billions of parameters with just a few lines of code.”

Impact Of AI In Human Life

Influencers also talked about the impact of artificial intelligence (AI) in our life and how it will change our future.

Source: Pixabay

Terence Mills retweeted an article titled Will A.I. Guarantees Our Humane Futures? This publication describes “how AI is a transformative technology while highlighting the risks associated with it.” In particular it explains how our lives are being changed by different AI-based smart algorithms that allow us to: get recommendations on our phones and websites, or use GPS-based navigation, or get a predictive text that completes the sentences when we try to send an e-mail or a message.

It also explains that “AI Engines and AI applications have already been adopted in almost every sector of the economy” helping to boost the countries’ growth and economy.

However, some downsides of AI are also presented via clear threads such as: digital exclusion, decrease employment in low-income jobs, violation of privacy and security rules, enforcement of bias and prejudices in the society.

Terence Mills also shared an article explaining that “AI is booming in the customer support industry” and how it enhances it, and why the introduction of “cognitive agents that teach themselves will change everything.” This post explains that AI agents with the ability to learn from humans on the fly and adjust their predictions, e.g. the answers from chatbots, can significantly increase the customer experience.

To reinforce this idea of AI being a game changer for multiple domains and sectors, DeepLearn007 has shared an article titled “How Will Human Life Change With the Advent of AIoT?”. It explains in particular “how AI combined with IoT will improve operations, make human-machine interaction better, and provide intelligent data management.”

Covid-19 Vaccination Program In The US

This week a number of influencers shared with their followers updates on the COVID19 vaccination programs in the US and tried to create awareness.

A 4-slide powerpoint presentation on the importance of getting vaccination was shared by Bob E. Hayes and Yann LeCunYann LeCun also retweeted the statement of the CDC Director Dr. Rochelle Walensky the data from the CDC suggest that vaccinated people do not carry the virus.

Jenn Wortman Vaughan tweeted that with the vaccination, she was able to meet her parents after one year.

Finally, Hilary Parker tweeted a tip to prevent muscle soreness for the people getting vaccinated.