HumanBeam is positioning “embodied AI” as a step beyond a text chatbot: a lifelike avatar trained on a defined knowledge base, delivered through a BeamBox-style 3D kiosk so the interaction feels like speaking with a front-desk companion. The emphasis is on agentic behavior—answering questions while also driving the next action (directions, check-in steps, forms, escalation), rather than dumping info and leaving the user to assemble the workflow. https://humanbeam.io/talktomeai
The demo was filmed at CES Las Vegas 2026 on the faytech booth, where the avatar runs on public-space display hardware instead of a typical monitor. faytech frames the install around an 86-inch 3840×2160 panel with optical bonding and infrared multi-touch, aimed at readability and durability for lobbies, clinics, and city kiosks. In the booth setup, they call out high-brightness operation (around the 1000-nit class) so the face and UI stay legible under show-floor lighting, in 4K.
For hospitality, the avatar becomes a travel guide or concierge trained on resort and local content, designed for walk-up, high-volume conversation and multilingual coverage (they cite 27 languages). When requests cross policy, liability, or “needs a human” boundaries, the same channel can switch from AI to a live remote staff member via a beam-in handoff, keeping context and reducing friction for late-night arrivals or accessibility needs, in care.
In education and healthcare, HumanBeam highlights virtual patient simulation for universities: configurable personas that let schools run repeatable ER and intake scenarios while observing how students ask questions and make decisions. On the operational side, the same interface can offload intake, wayfinding, and routine FAQ in a clinic, then escalate to a nurse or doctor only when needed—shifting humans away from admin loops and back toward empathy and triage.
A notable technical thread is “intent-based” interaction: the avatar infers what a visitor is trying to accomplish, captures qualified leads, and can surface context-relevant prompts without forcing a rigid script. The booth also acknowledges constraints such as needing reliable connectivity for some sessions, plus privacy and consent questions that come with vision cues, sentiment signals, and analytics in a public kiosk. The positioning is less “replace staff” and more “extend staff capacity” with a consistent, human-like front-end role.
I’m publishing about 100+ videos from CES 2026, I upload about 4 videos per day at 5AM/11AM/5PM/11PM CET/EST. Check out all my CES 2026 videos in my playlist here: https://www.youtube.com/playlist?list=PL7xXqJFxvYvjaMwKMgLb6ja_yZuano19e
This video was filmed using the DJI Pocket 3 ($669 at https://amzn.to/4aMpKIC using the dual wireless DJI Mic 2 microphones with the DJI lapel microphone https://amzn.to/3XIj3l8 ), watch all my DJI Pocket 3 videos here https://www.youtube.com/playlist?list=PL7xXqJFxvYvhDlWIAxm_pR9dp7ArSkhKK
Click the “Super Thanks” button below the video to send a highlighted comment under the video! Brands I film are welcome to support my work in this way 😁
Check out my video with Daylight Computer about their revolutionary Sunlight Readable Transflective LCD Display for Healthy Learning: https://www.youtube.com/watch?v=U98RuxkFDYY
