Google's AI Solve For Traffic, $249 Smart Glasses, Meta Flags Deepfakes

Halo, founded by former Harvard students AnhPhu Nguyen and Caine Ardayfio, is launching Halo X smart glasses that continuously listen, record, transcribe, and surface real-time prompts to the wearer. The $249 glasses open for preorder Wednesday and are backed by $1 million from Pillar VC, Soma Capital, Village Global, and Morningside Venture.

The frames have no external light to signal recording and offload processing to a phone app that pipes audio through Soniox transcription plus Google Gemini and Perplexity chatbots. Halo says each audio file is deleted after transcription, pledges future end-to-end encryption, and is pursuing SOC 2 compliance yet provides no timeline.

Privacy experts warn the device erodes conversational expectations and could violate two-party consent laws in over a dozen U.S. states, pushing liability onto users. The lack of a visible indicator differentiates Halo from Meta’s Ray-Ban Glasses and positions the startup squarely in regulatory and reputational crosshairs.

Amy Klobuchar turns a viral AI deepfake of herself into a rallying cry for Congress to set strict rules on manipulated content. The senator says a bipartisan NO FAKES Act would let Americans demand platforms remove unauthorized replicas of their voices and faces.

The fake clip, which drew more than a million views on X, stayed up while TikTok deleted it and Meta slapped an AI label. Klobuchar cites the recently signed TAKE IT DOWN Act as progress but calls it only a first step toward broader safeguards.

She warns deepfakes already tarnish reputations and could trigger crises faster than facts can counter them. Europe is moving faster, with the EU’s AI Act mandating clear labels and Denmark weighing full copyright over citizens’ faces and voices.

Professionals are sinking hours and money into AI apps that claim to enhance efficiency but often derail real work. Small, repeated credit purchases add up quickly, leaving pivotal projects unfinished.

Vendors use casino-style tactics such as low entry fees, unpredictable rewards, and credit systems to keep users experimenting. OpenAI and MIT research cited in the article finds heavy ChatGPT users show withdrawal symptoms and neglect relationships and tasks.

Companies facing multiple employees testing different AI tools see costs escalate into four-figure sums with little to show for it. The article concludes that constant “tool hopping” creates pseudo-productivity, turning the AI promise into a hidden expense line.

Bill Gates and the Alzheimer’s Disease Data Initiative unveiled a $1 million competition to apply agentic A.I. to Alzheimer’s and related dementia research. The winning team’s tools will be released publicly through the Initiative’s online environment.

Applications open today, with semi-finalists announced in December and finalists facing off next March at the Alzheimer’s Disease and Parkinson’s Disease Conference in Copenhagen. Gates Ventures funds the effort, building on Gates’s prior $130 million personal contributions toward treatments, trials, and diagnostics.

Organizers say agentic A.I. can autonomously scan vast biomedical datasets to surface biomarkers, refine trial design, and reveal drug targets faster than traditional methods. They stress the urgency as more than 7 million Americans already live with Alzheimer’s, a figure expected to rise.

Quebec City has begun synchronizing traffic lights using Google’s Green Light AI system. The partnership makes it the first municipality in Canada to deploy the tool, which Google provides at no cost.

The model analyzes anonymized Google Maps driving patterns and suggests precise timing tweaks engineers can implement within minutes. Initial changes at 11 intersections, including a 15-second gap reduction on a busy corridor, already show smoother rush-hour flow.

Across the 19 cities already using Green Light, Google reports up to 30% fewer stop-starts and a 10% cut in estimated intersection CO2 emissions. Quebec officials add that streamlined signals speed buses too, countering fears of increased car reliance.

Join the free, virtual AI:ROI Conference on September 25, 2025 to learn how to turn AI spend into measurable returns. Walk away with practical, no‑BS playbooks you can put to work next quarter.

Whether you’re a CEO, CFO, COO, or an AI champion building momentum, you’ll leave with templates, benchmarks, and case studies you can tailor to your industry, team size, and budget.

Hear from Scott Galloway, Mo Gawdat, and leaders from Moderna, Mars, IBM, Adobe, Logitech, and Horizon Media. Early adopters report 92% payback and up to 4x returns in about 13 months—join 20,000 AI innovators to see how.