Trending this week: Papers with code newsletter #4; 10 must-try open-source tools for machine learning; The best ML frameworks & extensions for Scikit-learn; The AI risks we should be focusing on.
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.
This week, Data Science and AI influencers on Twitter have talked about:
- ML Updates
- AI Threats
- Impact Of AI in Human Lives
The following sections provide all the details for each topic.
This week, some influencers have shared updates on machine learning and deep learning.
- Trending papers with code: the latest progress in self-attention vision models; Capsule network with self-attention routing; Unified framework for vision-and-language learning.
- Trending libraries/tools of the week: ZeRO-Offload, PixelLib, WeNet, and FinRL.
- Trending datasets of the week: ArtEmis, ToTTo, and DAF:re.
Ipfconline has shared an article presenting the 10 Must-Try Open Source Tools for Machine Learning. This post provides a list of open-source resources for:
- Non-programmers: Knime;Uber Ludwig.
- Model-deployment: TensorFlow.js;MLFlow.
- NLP, computer vision and audio: Detectron; SimpleCV; Tesseract OCR; Open AI Gym; Unity ML Agents.
- Data mining: Weka.
KDnuggets has shared an article presenting The Best Machine Learning Frameworks & Extensions for Scikit-learn. This post provides a list of popular tools and libraries that extend the Scikit-learn ecosystem. This list comprises:
- Libraries that can be used to process and transform data;
- tools and libraries that integrate Sklearn for better Auto-ML;
- Tools that you can use for machine learning experimentation;
- Libraries that are focused on model selection and optimization;
- Tools that you can use to export your models for production;
- Libraries that can be used for model visualization and inspection;
- Other complementary tools.
This week, some influencers have shared some content talking about AI threads, in particular those related to ethics.
Tamara McCleary has shared the two articles that follow:
An article that talks about the AI risks we should be focusing on. This post focus on the three following threats that AI can raise:
- Deepfakes can sow doubt and discord: produced content can be used to ruin reputations and commit fraud-based crimes. Deepfakes can fool viewers into believing fabricated statements or events are real;
- Large language models as disinformation force multipliers: for example, GPT-3 can produce sexist, racist, and discriminatory text because it learns from the internet content it was trained on;
- Ethics: we should go towards an ethical and socially beneficial AI. AI itself is neither inherently good nor bad. There are many real potential benefits that it can unlock for society — we just need to be thoughtful and responsible in how we develop and deploy it.
Another article titled “Ethical AI Is Our Responsibility”. This post talks about the ethical issues in AI. It describes the major causes of ethical AI breakdowns that it distills down into lack of sufficient: awareness, budget, time, and governance. This post also presents some key principles to consider to adhere to responsible AI. It considers that to be responsible AI should be: fair and unbiased, reliable and safe, inclusive, and private and secure.
KDnuggets has shared an article titled “DeepMind AGI paper adds urgency to ethical AI”. This post explains how reinforcement learning will one day lead to replicating human cognitive capabilities and achieve artificial general intelligence (AGI), and it states that we are not ready for it. Indeed, the potential for unintended consequences from AGI appears astronomical as it is difficult to enable transparency, eliminate bias, and ensure the ethical use of today’s narrow AI. There is the risk that businesses will prioritize profits and governments continue to surveil and control their populations. Also, AI concentrates power, with a relatively small number of companies controlling the technology.
Impact Of AI On Human Lives
This week a lot of data science influencers shared the latest methods by which AI is impacting human lives.
Ronald van Loon shared an infographic on the power of AI & Robotics in healthcare. The infographic describes how robotics is transforming the healthcare domain.
Yann LeCun shared an article that shows how AI is changing the way we shop. Advancing AI to make shopping easier for everyone is the main theme covered in this article. The article also talks about how the billion people who visit the Marketplace each month get the most relevant, state-of-the-art results when searching for products.
And finally, Terence Mills shared an article on how a Startup Adapts AI Used In Space To Advance Healthcare On Earth. The article talks about how startups are accelerating innovation with the cloud and they are bringing space technology down to Earth.