Modified protein-design tool could make it easier to tackle challenging drug targets — but AI antibodies are still a long way from reaching the clinic.
“They remove some of the magic,” said Dimitris Papailiopoulos, a machine learning researcher at the University of Wisconsin, Madison. “That’s a good thing.”
Training Transformers
Large language models are built around mathematical structures called artificial neural networks. The many “neurons” inside these networks perform simple mathematical operations on long strings of numbers representing individual words, transmuting each word that passes through the network into another. The details of this mathematical alchemy depend on another set of numbers called the network’s parameters, which quantify the strength of the connections between neurons.
I think AI agent workflows will drive massive AI progress this year — perhaps even more than the next generation of foundation models. This is an important trend, and I urge everyone who works in AI to pay attention to it.
CNBC’s Steve Kovach joins ‘Halftime Report’ to discuss the latest news on Microsoft’s new AI PC launch.
Machine-learning system trained on millions of human audio clips shows promise for detecting COVID-19 and tuberculosis.
This robot is equipped with AI-backed deep learning algorithms to autonomously manage assisting users with underlying physiological conditions.
The robot illustrated seamless functioning that supports users in walking, standing, and climbing stairs or ramps. Scientists call it, a “unified control framework.”
Apple quietly submitted a research paper last week related to its work on a multimodal large language model (MLLM) called MM1. Apple doesn’t explain what the meaning behind the name is, but it’s possible it could stand for MultiModal 1.
Being multimodal, MM1 is capable of working with both text and images. Overall, its capabilities and design are similar to the likes of Google’s Gemini or Meta’s open-source LLM Llama 2.
An earlier report from Bloomberg said Apple was interested in incorporating Google’s Gemini AI engine into the iPhone. The two companies are reportedly still in talks to let Apple license Gemini to power some of the generative AI features coming to iOS 18.
Google just announced that it has been riverline floods, up to seven days in advance in some cases. This isn’t just tech company hyperbole, as the findings were actually published Nature. Floods are the most common natural disaster throughout the world, so any early warning system is good news.
Floods have been notoriously tricky to predict, as most rivers don’t have streamflow gauges. Google got around this problem by with all kinds of relevant data, including historical events, river level readings, elevation and terrain readings and more. After that, the company generated localized maps and ran “hundreds of thousands” of simulations in each location. This combination of techniques allowed the models to accurately predict upcoming floods.
Particles colliding in accelerators produce numerous cascades of secondary particles. The electronics processing the signals avalanching in from the detectors then have a fraction of a second in which to assess whether an event is of sufficient interest to save it for later analysis. In the near future, this demanding task may be carried out using algorithms based on AI, the development of which involves scientists from the Institute of Nuclear Physics of the PAS.
While Elon Musk says Tesla is trying to build an AI supercomputer, his companies are spending billions of dollars on Nvidia hardware.
