Combining Learning and Reasoning, Physics from Videos, and the EU AI Act: AIhub Monthly Digest for June 2023
Introduction:
Welcome to our June 2023 monthly digest! Here at AIhub, we aim to keep you informed about all the latest developments in the world of artificial intelligence. In this edition, we have some fascinating stories to share, including the use of deep learning and automated reasoning to solve complex problems, learning physics from videos, and scalable and robust planning in lifelong reinforcement learning. We also have updates on recent events such as the VISION AI Open Day and the winners of the IJCAI awards. Plus, we take a look at the progress towards EU AI regulation and the potential risks of large language models. Don’t miss our machine learning course for beginners and a fun video explaining AI in just 4 minutes. Stay tuned for more exciting updates!
Full Article: Combining Learning and Reasoning, Physics from Videos, and the EU AI Act: AIhub Monthly Digest for June 2023
June 2023 Monthly Digest: AIhub Stories, News, Events, and More
Welcome to our June 2023 monthly digest, where you can catch up on any AIhub stories you may have missed, peruse the latest news, find out about recent events, and more. In this edition, we explore the combination of deep learning and automated reasoning to solve complex problems, learn physics from videos, and congratulate the IJCAI award winners.
Bridging the Gap Between Learning and Reasoning
Solving Sudoku puzzles and designing proteins may seem unrelated, but Marianne Defresne sheds light on their connection in her blog post. She discusses the work of Sophie Barbe and Thomas Schiex, who have successfully combined deep learning with automated reasoning to tackle complex problems.
Interview with Haotian Xue: Learning Intuitive Physics from Videos
Haotian Xue, along with Antonio Torralba, Joshua Tenenbaum, Daniel Yamins, Yunzhu Li, and Hsiao-Yu Tung, presents a framework for learning 3D-grounded visual intuitive physics models from videos of complex scenes with fluids. In an interview, Haotian shares insights into their work and methodology.
Scalable and Robust Planning in Lifelong Reinforcement Learning
Safa Alver and Doina Precup introduce models that enable scalable and robust planning in lifelong reinforcement learning scenarios. Their paper, “Minimal Value-Equivalent Partial Models for Scalable and Robust Planning in Lifelong Reinforcement Learning,” delves into the details of these models.
VISION AI Open Day: Trustworthy AI
The second VISION AI Open Day, focused on “trustworthy AI,” was held in Prague. The event featured a live-streamed roundtable discussion, exploring various aspects of trustworthy AI. Find out more about this event and watch the panel discussion by visiting the provided link.
IJCAI 2023 Awards Announced
Three IJCAI awards have been announced to commemorate exceptional contributions to the field. The winners include Sarit Kraus for the Award for Research Excellence, Dieter Fox for the John McCarthy Award, and Pin-Yu Chen for the Computers and Thought Award. Learn more about these awards on the IJCAI webpage.
Next Step Taken Towards EU AI Regulation
On June 14, the European Parliament voted in favor of adopting EU AI legislation. With 499 votes in favor, 28 against, and 93 abstentions, this marks an important milestone in the process. The next step involves talks with EU member states to finalize the law’s form. For more information, refer to the provided document.
Science in the Age of Large Language Models
Abeba Birhane, Atoosa Kasirzadeh, David Leslie, and Sandra Wachter delve into the topic of large language models and their associated risks in their paper, “Science in the Age of Large Language Models.” They emphasize the need for careful consideration and responsible usage to preserve good scientific practices and trust in science. The authors’ video discussion on this topic is also available.
Report on LLMs for Non-English Content
The Center for Democracy and Technology (CDT) presents a report analyzing new language models capable of analyzing text across languages. Authors Gabriel Nicholas and Aliya Bhatia explore these models’ capabilities and limits, providing valuable insights into the growing field.
Machine Learning for Beginners: An Online Course
Microsoft offers a 26-lesson online course called “ML for Beginners,” targeting individuals new to machine learning. Additionally, a video series accompanies the course, guiding students through selected lessons and offering practical exercises and quizzes. The course covers clustering, classification, reinforcement learning, and more.
AI Explained in 4 Minutes
If you’re new to artificial intelligence, Cyber Simplicity has an informative YouTube video explaining the basics. The video adopts the style of a British 1960’s public information film, making it both entertaining and educational.
Stay Informed with AIhub
To further expand your knowledge and stay up-to-date, visit our resources page and explore forthcoming and past seminars for 2023. We also feature focus series on AI around the world and the UN Sustainable Development Goals, as well as the New Voices in AI series.
Tags: monthly digest
Summary: Combining Learning and Reasoning, Physics from Videos, and the EU AI Act: AIhub Monthly Digest for June 2023
Welcome to our June 2023 monthly digest, where you can catch up on AIhub stories, read the latest news, and learn about recent events. In this month’s digest, we explore how deep learning and automated reasoning are being combined to solve complex problems. We also interview Haotian Xue about their work on learning intuitive physics from videos. Additionally, we delve into scalable and robust planning in lifelong reinforcement learning. Other highlights include the VISION AI Open Day on “trustworthy AI” and the announcement of the IJCAI 2023 awards. We also discuss the progress towards EU AI regulation and the potential risks of large language models. Plus, don’t miss Microsoft’s ML for beginners course and a quick explanation of AI in a retro-style video. Stay informed with AIhub!
Frequently Asked Questions:
Q1: What is Artificial Intelligence (AI)?
A1: Artificial Intelligence, commonly referred to as AI, is a branch of computer science that focuses on creating smart machines capable of performing tasks that typically require human intelligence. These tasks include speech recognition, decision-making, problem-solving, learning, and more.
Q2: How does Artificial Intelligence work?
A2: AI employs various techniques such as machine learning, natural language processing, and deep learning. Machine learning involves training algorithms on large datasets to enable systems to automatically learn and improve from experience. Natural language processing allows machines to understand and communicate with humans in a more human-like manner. Deep learning utilizes artificial neural networks to process data and learn patterns, making it particularly useful for complex tasks like image and speech recognition.
Q3: What are the applications of Artificial Intelligence?
A3: Artificial Intelligence has numerous applications across various industries. It is commonly used in autonomous vehicles, virtual assistants (such as Siri and Alexa), fraud detection systems, personalized recommendation systems, robotics, healthcare diagnosis, financial analysis, and many more. The potential uses of AI are vast and continue to expand as technology advances.
Q4: What are the ethical concerns surrounding Artificial Intelligence?
A4: As AI becomes more sophisticated, ethical concerns arise. One concern is the potential impact on the job market, as AI automation could replace certain jobs, leading to unemployment and socioeconomic inequality. Other ethical concerns include privacy issues, biases in AI decision-making, and the use of AI in military applications. Addressing these concerns and implementing regulations are crucial to ensure AI is used responsibly and for the greater good.
Q5: What are the future prospects of Artificial Intelligence?
A5: The future prospects of AI are incredibly promising. Advancements in AI technology will likely lead to significant improvements in various fields, including healthcare, transportation, education, and customer service. AI has the potential to revolutionize industries, enhance efficiency, and improve the overall quality of life. However, it is important to carefully navigate its development and ensure ethical considerations are taken into account to reap the benefits without compromising human well-being.
