Lifeboat Foundation: Safeguarding Humanity
Toggle light / dark theme

Blog

Category: robotics/AI – Page 819

The Kessler Effect and how to stop it

NASA space debris expert Don Kessler observed that, once past a certain critical mass, the total amount of space debris will keep on increasing: collisions give rise to more debris and lead to more collisions, in a chain reaction. So Clean Space is seeking not just to cut debris production from future ESA missions but to reduce the total mass of current debris, such as the robotic salvage of derelict satellites. The task is an urgent one: debris levels have increased 50% in the last five years in low orbit.

AI aids colonoscopy by acting as ‘extra pair of eyes’, spotting tumors

He further mentioned that the AI tool functioned like an “extra pair of eyes,” identifying potential tumors within the video footage.

In short, the AI tool assists junior doctors during colonoscopies by analyzing video footage from the endoscope and identifying potential tumors. It aids in detecting adenomas, particularly those smaller than five millimeters (mm) in diameter.

Scientists fuse brain-like tissue with electronics to make computer

Scientists have fused brain-like tissue with electronics to make an ‘organoid neural network’ that can recognise voices and solve a complex mathematical problem. Their invention extends neuromorphic computing – the practice of modelling computers after the human brain – to a new level by directly including brain tissue in a computer.

The system was developed by a team of researchers from Indiana University, Bloomington; the University of Cincinnati and Cincinnati Children’s Hospital Medical Centre, Cincinnati; and the University of Florida, Gainesville. Their findings were published on December 11.

Meet Netron: A Visualizer for Neural Network, Deep Learning and Machine Learning Models

Exploring pre-trained models for research often poses a challenge in Machine Learning (ML) and Deep Learning (DL). Visualizing the architecture of these models usually demands setting up the specific framework they were trained on, which can be quite laborious. Without this framework, comprehending the model’s structure becomes cumbersome for AI researchers.

Some solutions enable model visualization but involve setting up the entire framework for training the model. This process can be time-consuming and intricate, deterring quick access to model architectures.

One solution to simplify the visualization of ML/DL models is the open-source tool called Netron. This tool functions as a viewer specifically designed for neural networks, supporting frameworks like TensorFlow Lite, ONNX, Caffe, Keras, etc. Netron bypasses the need to set up individual frameworks by directly presenting the model architecture, making it accessible and convenient for researchers.

New insight into how brain adjusts synaptic connections during learning may inspire more robust AI

How the brain adjusts connections between #neurons during learning: this new insight may guide further research on learning in brain networks and may inspire faster and more robust learning #algorithms in #artificialintelligence.

Researchers from the MRC Brain Network Dynamics Unit and Oxford University’s Department of Computer Science have set out a new principle to explain how the brain adjusts connections between neurons during learning. This new insight may guide further research on learning in brain networks and may inspire faster and more robust learning algorithms in artificial intelligence.

The essence of learning is to pinpoint which components in the information-processing pipeline are responsible for an error in output. In , this is achieved by backpropagation: adjusting a model’s parameters to reduce the error in the output. Many researchers believe that the brain employs a similar learning principle.

However, the biological brain is superior to current machine learning systems. For example, we can learn new information by just seeing it once, while artificial systems need to be trained hundreds of times with the same pieces of information to learn them. Furthermore, we can learn new information while maintaining the knowledge we already have, while learning new information in artificial neural networks often interferes with existing knowledge and degrades it rapidly.

In 2024 AI will make it almost impossible to know the truth

All these previous innovations pale in the face of tools like MidJourney, DALL-E, and Adobe Firefly.

These generative AI image systems, the kind that easily spits out this image below of a flooded downtown Manhattan, are dream weavers that make the literal out of the imagined.

When Midjourney builds an image, there are no easily identifiable sources, mediums, or artists. Every pixel can look as imaginary or real as you want and when they leave the digital factory, these images (and video) travel fleetfooted around the world, leaving truth waiting somewhere in the wilderness.

Samsung unveils its first robot vacuum-mop that uses AI to detect stains and can steam clean floors

Samsung is planning to release what might be the most advanced cleaning robot yet: a robot vacuum and mopper that will steam clean floors and use AI to detect stains.

The upcoming “Bespoke Jet Bot Combo” cleaner will have a charging base that will auto wash, clean, and dry the robot’s mop pads.

The device will also use AI to detect floor types, objects, and spot stains. When the robot detects a stain, “it goes back to the clean station to heat the mop pads with high-temperature steam and water and then returns to the area,” says a press release on Samsung’s website.

/* */