Brain implant user shows how speech BCIs are moving into daily life

Casey Harrell, who has ALS, is using a brain-computer interface to communicate, work and browse the web. South Korea’s enthusiasm for AI also stands out as public concern rises elsewhere.

Casey Harrell, who has ALS and is paralyzed, has had electrodes embedded in his brain for almost three years and first used his brain-computer interface to “speak” in 2023. Since then, he has logged thousands of hours with the system, using it largely independently to communicate, browse the web and perform his job.

The team behind the device describes Harrell as “the first power user of a speech BCI” and is planning more enhancements. Harrell framed the technology as a way to preserve ambition while living with ALS, saying diminished dreams should not be assumed for people with the disease.

South Korea’s unusually positive view of AI contrasts with growing skepticism in the US. Only 16% say they are more concerned than excited about AI, the lowest share among the 25 countries surveyed by the Pew Research Center, while 50% of Americans were more worried than excited. The country’s outlook is tied to a broad belief that adopting technology is essential to modernization and global influence.

Other technology developments include US restrictions on Anthropic AI over foreign intelligence risks, Alibaba’s release of AI models for robots, EA’s push to place advertising directly into gameplay, reports that Reddit can be used to manipulate AI search, and new concerns over deepfake detection and Meta’s AI reorganization.

62

Impact Score

AMD and Rackspace plan dedicated AI compute rollout

AMD and Rackspace have finalized a phased deployment for dedicated AMD-based compute across Rackspace data centers. The capacity is aimed at regulated enterprise workloads, including clinical AI and large-scale inference.

Lexar tests SSD offloading for local AI models

Lexar is developing an AI-focused SSD approach designed to cut DRAM demand when running large language models on consumer PCs. Internal tests show the company’s storage offloading can load models that traditional local frameworks struggle to run with limited memory.

NVIDIA Blackwell leads MLPerf Training 6.0

NVIDIA’s latest MLPerf Training 6.0 results put Blackwell across every benchmark in the suite, including new MoE workloads. Partner systems from Microsoft Azure and CoreWeave highlighted large-cluster runs on Llama 3.1 405B and DeepSeek-V3 671B.

HPE and NVIDIA expand AI Factory for agentic systems

HPE and NVIDIA are adding agent tooling, confidential computing and updated accelerated systems across the HPE AI Factory portfolio. The expansion targets production deployments that need governance, secure data handling and integrated networking.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.