Lifeboat Foundation

Deep Learning (DL) performs classification tasks using a series of layers. To effectively execute these tasks, local decisions are performed progressively along the layers. But can we perform an all-encompassing decision by choosing the most influential path to the output rather than performing these decisions locally?

In an article published today in Scientific Reports, researchers from Bar-Ilan University in Israel answer this question with a resounding “yes.” Pre-existing deep architectures have been improved by updating the most influential paths to the output.

Continue reading “Better paths yield better AI: Enhancing pre-existing architectures” »

There’s a proverb in astronomy that goes something like, “black holes have no hair.” This indicates that black holes are extremely straightforward entities under the framework of general relativity. The only necessary characteristics of a black hole are its mass, electric charge, and spin rate. You now know everything there is to know about black holes just from those three numbers. That is to say, they are bare; they lack any further data.

This feature of black holes has been a major source of frustration for astronomers trying to figure out the inner workings of these cosmic behemoths. However, understanding black holes and their inner workings is impossible due to the absence of any kind of “hair” on their surfaces. Unfortunately, black holes continue to be among the universe’s most elusive and baffling features.

The present knowledge of general relativity, however, is essential to the “no-hair” black hole notion. The emphasis of this relativity illustration is on the curved nature of space-time. Any object with enough mass or energy to bend space-time around it will provide that object directions for movement.

Join Dr. Ben Goertzel, the visionary CEO and Founder of SingularityNET, as he delves into the compelling realm of large language models. In this Dublin Tech Summit keynote presentation, Dr. Goertzel will navigate the uncharted territories of AI, discussing the imminent impact of large language models on innovation across industries. Discover the intricacies, challenges, and prospects of developing and deploying these transformative tools. Gain insights into the future of AI, as Dr. Goertzel unveils his visionary perspective on the role of large language models in shaping the AI landscape. Tune in to explore the boundless potentials of AI and machine learning in this thought-provoking session.

Themes: AI & Machine Learning | Innovation | Future of Technology | Language Models | Industry Transformation.
Keynote: Dr. Ben Goertzel, CEO and Founder, SingularityNET
#dubtechsummit

Summary: The mushroom body—a key area in the brains of arthropods like insects—plays a crucial role in abstract behavioral decision-making.

Contrary to the long-standing belief that insects react purely on stimulus-response, the study shows they can actually make nuanced decisions based on experiences. The researchers recorded feeding behavior alongside neural signals.

This has implications for understanding not just insect behavior but also basic neurobiological principles that are similar in humans.

An ape skull found in Turkey may challenge the belief that human and ape ancestors came from Africa. The discovery suggests that hominins may have first evolved.

Robots based on soft materials are often better at replicating the appearance, movements and abilities of both humans and animals. While there are now countless soft robots, many of these are difficult to produce on a large-scale, due to the high cost of their components or their complex fabrication process.

Researchers at University of Coimbra in Portugal recently developed a new soft robotic hand that could be more affordable and easier to fabricate. Their design, introduced in Cyborg and Bionic Systems, integrates soft actuators with an exoskeleton, both of which can be produced using scalable techniques.

Continue reading “Scientists create soft and scalable robotic hand based on multiple materials” »

Large language models (LLMs) have become a general-purpose approach to embodied artificial intelligence problem-solving. When agents need to understand the semantic nuances of their environment for efficient control, LLMs’ reasoning skills are crucial in embodied AI. Recent methods, which they refer to as “programs of thought,” use programming languages as an improved prompting system for challenging reasoning tasks. Program-of-thought prompting separates the issues into executable code segments and deals with them one at a time, unlike chain-of-thought prompting. However, the relationship between the use of programming languages and the development of LLMs’ thinking skills has yet to receive enough research. When does program-of-thought suggesting work for reasoning2 remain the crucial question?

The complexity-impacted reasoning score (CIRS), a thorough metric for the link between code reasoning stages and their effects on LLMs’ reasoning abilities, is proposed in this paper. They contend that programming languages are inherently superior to serialized natural language because of their improved modeling of complex structures. Their innate procedure-oriented logic aids in solving difficulties involving several steps in thinking. Because of this, their suggested measure assesses the code complexity from both a structural and a logical standpoint. In particular, they compute the structural complexity of code reasoning stages (rationales) using an abstract syntax tree (AST). Their method uses three AST indicators (node count, node type, and depth) to keep all structural information in AST represented as a tree, which thoroughly comprehends code structures.

Researchers from Zhejiang University, Donghai Laboratory and National University of Singapore develop a way to determine logical complexity by combining coding difficulty with cyclomatic complexity, drawing inspiration from Halsted and McCabe’s idea. Thus, it is possible to consider the code’s operators, operands, and control flow. They can explicitly calculate the logic’s complexity within the code. They discover through an empirical investigation using their suggested CIRS that present LLMs have a restricted comprehension of symbolic information like code and that not all sophisticated code data can be taught and understood by LLMs. Low-complexity code blocks lack the necessary information, but high-complexity code blocks could be too challenging for LLMs to understand. To effectively improve the reasoning abilities of LLMs, only code data with an appropriate amount of complexity (structure & logic), both basic and detailed, are needed.

Incidentally, a few days ago I received a message from my paleobiologist colleague Dr. Ken Towe, a retired senior scientist at the Smithsonian Institution.

In a major breakthrough, scientists have built a tool to predict the odor profile of a molecule, just based on its structure. It can identify molecules that look different but smell the same, as well as molecules that look very similar but smell totally different. The research was published in Science.

Professor Jane Parker, University of Reading, said, “Vision research has wavelength, hearing research has frequency—both can be measured and assessed by instruments. But what about smell? We don’t currently have a way to measure or accurately predict the odor of a molecule, based on its molecular structure.”

“You can get so far with current knowledge of the molecular structure, but eventually you are faced with numerous exceptions where the odor and structure don’t match. This is what has stumped previous models of olfaction. The fantastic thing about this new ML generated model is that it correctly predicts the odor of those exceptions.”