Lifeboat Foundation

We are in the middle of a data-driven science boom. Huge, complex data sets, often with large numbers of individually measured and annotated ‘features’, are fodder for voracious artificial intelligence (AI) and machine-learning systems, with details of new applications being published almost daily.

But publication in itself is not synonymous with factuality. Just because a paper, method or data set is published does not mean that it is correct and free from mistakes. Without checking for accuracy and validity before using these resources, scientists will surely encounter errors. In fact, they already have.

In the past few months, members of our bioinformatics and systems-biology laboratory have reviewed state-of-the-art machine-learning methods for predicting the metabolic pathways that metabolites belong to, on the basis of the molecules’ chemical structures¹. We wanted to find, implement and potentially improve the best methods for identifying how metabolic pathways are perturbed under different conditions: for instance, in diseased versus normal tissues.

Window to the soul? Maybe, but the eyes are also a flashing neon sign for a new artificial intelligence-based system that can read them to predict what you’ll do next.

A University of Maryland researcher and two colleagues have used eye-tracking technology and a new deep-learning AI algorithm to predict study participants’ choices while they viewed a comparison website with rows and columns of products and their features.

The algorithm, known as RETINA (Raw Eye Tracking and Image Ncoder Architecture), could accurately zero in on selections before people had even made their decisions.

GeoMindGPT, a customized version of ChatGPT, powered by GPT-4, is p ioneering the frontier of AI-assisted understanding of complex scientific and philosophical concepts with a special focus on Global Superintelligence, Technological Singularity, Transhumanism & Posthumanism, Consciousness Studies, Quantum Gravity, Simulation Metaphysics.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 The world’s largest music label has yanked its artists’ music off TikTok Universal Music Group claims TikTok is unwilling to compensate musicians appropriately. (The Guardian) + Taylor Swift fans are kicking off. (Wired $) + Indie record labels don’t like the sound of Apple’s pay plans either. (FT $)

Researchers found Llama 2 fared not much better than random guessing in a medical test, while GPT-4 almost got a passing grade.

Researchers have overcome a major challenge in biomimetic robotics by developing a sensor that, assisted by AI, can slide over braille text, accurately reading it at twice human speed. The tech could be incorporated into robot hands and prosthetics, providing fingertip sensitivity comparable to humans.

Human fingertips are incredibly sensitive. They can communicate details of an object as small as about half the width of a human hair, discern subtle differences in surface textures, and apply the right amount of force to grip an egg or a 20-lb (9 kg) bag of dog food without slipping.

Continue reading “Breakthrough could see robots with ‘fingertips’ as sensitive as humans” »

We’re on a journey to advance and democratize artificial intelligence through open source and open science.

Researchers mapped over 10,000 mouse hippocampal #neurons, creating the world’s most comprehensive database of single-neuron #connectivity #patterns.

Summary: Researchers unveiled the most extensive single-neuron projectome database to date, featuring over 10,000 mouse hippocampal neurons.

The study provides an unprecedented view of the spatial connectivity patterns at the mesoscopic level, crucial for understanding learning, memory, and emotional processing in the hippocampus. By employing machine learning algorithms for categorizing axonal trajectories and integrating spatial transcriptome data, researchers identified 43 distinct projectome cell types, revealing intricate projection patterns and soma locations’ correspondence to projection targets.

Continue reading “Mapping the Brain: The Largest Neuron Projectome Unveiled” »

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Human babies are far better at learning than even the very best large language models.