Meta Platforms Inc. is building a new gigawatt-sized data center in Texas to advance its artificial intelligence efforts, the latest in a string of significant investments by the company as it looks to keep pace in the competitive AI industry.
A study led by Sylvester Comprehensive Cancer Center, part of University of Miami Miller School of Medicine (FL, USA) seeks to understand how AI can improve breast cancer screening. The Pragmatic Randomized Trial of Artificial Intelligence for Screening Mammography (PRISM) trial will examine hundreds of thousands of mammograms to “assess AI’s true impact”
Despite huge investments in research, breast cancer remains a leading cause of mortality in US women. Routine mammography has increased the diagnosis of early-stage cancer, but the increased incidence of false positives can lead to unnecessary testing, anxiety and higher costs.
“As the first major randomized trial of AI in breast cancer screening in the US, this study represents a pivotal step,” commented Jose Net, University of Miami Miller School of Medicine and co-principal investigator of the study. “Our goal is to rigorously and objectively assess AI’s impact, identifying who benefits and who may not.”
Discover how Artificial Intelligence in Cancer Drug Discovery accelerates target identification, drug design, biomarkers, and clinical trials.
The program seeks new kinds of cognitive radio techniques that enable wireless communications that autonomously find open radio frequencies and choose the most efficient RF waveform to avoid interference, achieve necessary range, and send data quickly.
Intelligent RF transceivers
Cognitive radio describes an RF transceiver that intelligently can detect which communication channels are in use, which ones are not, and instantly move into vacant channels. The same principles could apply to radar, electronic warfare (EW) and other RF and microwave applications.
Many of the environments where human-facing universal robots can provide benefits — homes, hospitals, schools — are sensitive and personal. A tutoring robot helping your kids with math should have a track record of safe and productive sessions. An elder-care assistant needs a verifiable history of respectful, competent service. A delivery robot approaching your front door should be as predictable and trustworthy as your favorite mail carrier. Without trust, adoption will never take place, or quickly stall.
Trust is built gradually and also reflects common understanding. We design our systems to be explainable: multiple AI modules talk to each other in plain language, and we log their thinking so humans can audit decisions. If a robot makes a mistake — drops the tomato instead of placing it on the counter — you should be able to ask why and get an answer you can understand.
Over time, as more robots connect and share skills, trust will depend on the network too. We learn from peers, and machines will learn from us and from other machines. That’s powerful but just like parents are concerned about what their kids learn on the web, we need good ways to audit and align skill exchange for robots… Governance for human–machine societies isn’t optional; it’s fundamental infrastructure.