Generative AI can’t understand the world
To be fair, who can?
MIT boffins have demonstrated that even the best-performing large language models (LLMs) lack a coherent understanding of the world and its rules, which can lead to unexpected failures on similar tasks.
MIT engineers can stick a solar panel on anything
Sticky
MIT boffins have created solar cells that can adhere to many different surfaces to provide power on the go.
MIT boffins come up with cheap fast charging batteries
Who needs Lithium?
Boffins at MIT describe new aluminum-sulfur batteries that are made entirely from abundant and inexpensive materials and can be charged in less than a minute.
MIT searches for cause of network jitters
Key algorithms to control "congestion" delays are "deeply unfair
A new study finds that many key algorithms designed to control "congestion" delays on computer networks "are prove deeply unfair, letting some users hog all the bandwidth while others get essentially nothing."
MIT speeds up hardware by mimicking computer brains
Well not yours on a Monday morning, obviously
MIT boffins have built AI chips that mimic synapses, but are a million times faster, and are more energy efficient than current designs.
MIT develop paper-thin speakers
The base is rubbish
MIT engineers have developed a paper-thin loudspeaker that they say turn any surface into an active audio source.
Boffins come up with better connections
Faster than USB or Thunderbolt
Boffins have emerged from their smoke-filled labs with a cabling system that could provide data transfer speeds multiple times faster than existing USB connections using an extremely thin polymer cable.
MIT boffins use AI to spot COVID-19
Sound of hacking
Scientists from MIT have developed a new AI model that can detect COVID-19 from a simple forced cough.
Stop teaching programming using logic puzzles!
Coding introduced to kids badly
MIT's Mitchel Resnick and Natalie Rusk have warned that trying to teach kids programming using logic puzzles is like teaching people to write using crosswords.
MIT kills racist AI database
It lowered the standards of the world
MIT has taken offline its highly cited dataset that trained AI systems to potentially describe people using racist, misogynistic, and other problematic terms.