Guy Hoffman
Most organisations are deploying AI into environments designed for people, then expecting the people to adapt. The result is friction that looks like a technology problem and is actually a collaboration problem: badly timed hand-offs, brittle trust, staff working around the system rather than with it. The buyers who feel this most acutely are the ones who have passed the pilot stage and are now trying to make human and machine teams productive at scale.
Guy Hoffman is a Cornell roboticist who helps organisations design AI and robotic systems that people can actually work with, not around.
Full Profile
Why organisations work with Guy Hoffman
- He runs one of the leading labs in the world on human-robot and human-AI collaboration, so the content is drawn from active research, not recycled commentary.
- His design training at Parsons sits on top of an MIT computer science doctorate, which is why his work on non-anthropomorphic robots and nonverbal machine communication reads as rigorous rather than aesthetic.
- His TEDxJaffa talk “Robots with Soul” passed three million views and remains one of the most-watched public explanations of why machines feel awkward to work with, and how to fix it.
- He has shipped work that reached wide audiences: the AUR robotic desk lamp won the IEEE Robot Design Competition, and his collaboration on the Digital Water Pavilion was named a TIME Best Invention of the Year.
- He speaks to leadership audiences about AI without the two failure modes buyers dread: pure hype on one side, and pure academic detachment on the other.
Biography highlights
- Associate Professor and Mills Family Faculty Fellow, Sibley School of Mechanical and Aerospace Engineering, Cornell University.
- Heads the Human-Robot Collaboration and Companionship (HRC2) group at Cornell.
- PhD, MIT Media Lab (human-robot interaction); MSc Computer Science, Tel Aviv University; animation studies, Parsons School of Design.
- Best Paper awards at HRI and robotics conferences across 2004, 2006, 2008, 2010, 2013, 2015, 2018, 2019, 2020 and 2021.
- Co-author of “Social Robot Morphology: Cultural Histories of Robot Design” in Cultural Robotics (Springer, 2023).
- Work covered by CNN, BBC, The New York Times, PBS, NBC, IEEE Spectrum and Semafor.
Biography
The hard problem in applied AI is not the model. It is the hand-off. The moment a machine has to time its action to a human’s, read context, share a task, or back off when the person wants to lead, most systems fail in ways that erode trust faster than any single output error. Guy Hoffman has spent two decades building the science of that moment.
At Cornell, Hoffman leads the Human-Robot Collaboration and Companionship group as Associate Professor and Mills Family Faculty Fellow in the Sibley School of Mechanical and Aerospace Engineering. The lab’s work sits across mechanical engineering, computer science and information science, which is why it produces research on timing, nonverbal communication and the design of machines that do not pretend to be human. His doctorate is from the MIT Media Lab; he also trained in animation at Parsons, and that combination shapes the questions he asks.
The public version of that thinking is his TEDxJaffa talk “Robots with Soul,” watched more than three million times and still one of the clearest explanations of why so many AI products feel wrong to use. The academic version is a body of work that has been recognised with Best Paper awards at HRI and robotics conferences across a decade and more, and a 2023 Springer chapter, “Social Robot Morphology,” that traces how cultural assumptions get baked into machine design.
For senior audiences, the through-line is practical. Companies are now deploying AI into workflows that were built for humans. Hoffman’s research shows what breaks at that boundary, why it breaks, and what good looks like when a machine is designed to collaborate rather than merely perform. That is a more specific brief than the usual AI keynote, and it lands with operators who have already moved past the demo.
Key speaking topics
- Human-AI collaboration and the design of productive machine partners
- Human-robot interaction in workplaces and homes
- Nonverbal communication between people and intelligent systems
- Non-anthropomorphic robot and AI design
- Timing, trust and hand-offs in human-machine teams
- Cultural assumptions embedded in AI and robot design
- The future of robots in care, service and creative work
Ideal for
- Chief technology and chief AI officers leading enterprise AI deployment
- Heads of R and D, innovation and product in technology-intensive sectors
- Boards and executive teams setting strategy for automation and AI-enabled services
- Health, manufacturing and service leaders introducing robotic or AI systems into human workflows
Audience outcomes
- A clearer map of where AI and robotic systems tend to fail at the human boundary, and why those failures are predictable
- A working vocabulary for discussing trust, timing and hand-offs in AI-enabled operations
- Specific design principles from two decades of human-robot interaction research, translated for non-technical leaders
- A view of where near-term robotics is heading in workplaces, homes and care settings, grounded in current lab work
- Sharper questions to bring back to internal AI and automation roadmaps
Talks
A live-demo keynote on why machines feel awkward to work with and what changes when designers borrow from animation, theatre and jazz.
Key takeaways:
- Why human reactions to AI are driven more by timing and gesture than by intelligence
- How non-anthropomorphic design can make machines feel more, not less, trustworthy
- What product and engineering teams can take from performance arts research
A talk on the governance and design choices that decide whether AI systems amplify human agency or quietly erode it.
Key takeaways:
- Where control is actually lost in human-AI systems, and where it is only assumed to be lost
- Design patterns that keep human judgement load-bearing inside automated workflows
- The leadership questions to ask before signing off on an AI deployment
A research-led session on embedding values into AI and robotic systems as a design discipline rather than a compliance exercise.
Key takeaways:
- How cultural assumptions get baked into machine behaviour
- Practical methods for surfacing value conflicts early in product design
- What “responsible AI” looks like when it is built in at the interaction layer