Our Mission

Founded in 2018, Sanctuary is an artificial general intelligence (AGI) company. Our mission is to build machines like us.

Building machines that think like people is incredibly hard. Here are some questions we have been asked about what we do.


WHAT IS YOUR business model?

In the short term, our business model is predicated on corporate partnerships.  Forward-thinking organizations that are keen to position themselves on the frontier of AI, robotics, and automation, are partnering with us on experiments designed to gain insight into how their customers will interact with empathic, human-like synths. 

We work with our partners to design experiments that help them learn the efficacy of certain types of human-machine interactions that are specific to their operations.  Specifically, we collaborate with our partners on identifying situations where empathy in a face-to-face situation is valuable and delivered at scale. Examples include elderly care and companionship, care in children’s hospitals, private education, hospitality, and retail including customer-facing sales and service. 

We then run the experiments in our Empathy Lab, sending our partners interim reports and videos so they can monitor progress and offer input along the way.  Finally, we prepare and present final reports with analytical results and video highlights reporting key findings in order to help our partners develop their longer-term AI strategies.

The typical duration of an experiment, from ideation to final presentation, is six months. We run experiments in parallel, so partners are able to test multiple hypotheses simultaneously.  We supplement our experiments with additional guidance on AI strategy development.

For more information, please contact us at partners@sanctuary.ai.



There is no good answer to this question. Even now, our world is undergoing dramatic and unpredictable changes that are caused by technological advances nowhere near the importance of AGI.

We can’t know how the transition to a world with AGI will happen. However, we are optimistic about a future where humans and synths can live together, and we strive to create situations where society overall benefits from both types of intelligent beings.



We are focused on building human-like intelligence, and embodying that intelligence in robots that are physically as human-like as we can make them. We view AGI as a robotics control system problem.

We want to build a control system that inputs the sensory data from the robot -- things like camera, microphone and touch sensor data -- and outputs action, which are signals sent to the robot's motors to move its body. A control system is more human-like -- and therefore better -- if the actions the robot takes are closer to what we'd expect a human to do, given the situation.


The basic premise that underlies our approach to understanding intelligence is that it is a property of a physical body, and cannot be separated from it. Intelligence evolved to help organisms better navigate their bodies through an environment and manipulate that environment to help the organism achieve its goals. We believe that in order for a mind to understand the world in the same way we do, it needs a body that is as close to ours as possible. This will allow it to explore through sight, sound, smell, taste and touch, which allows the AI to develop concepts and an understanding of the world that is similar to the way our own minds work.



The entirety of human civilization is the product of our minds and the way they work. Every future advance, from curing cancer and aging, to building quantum computers, to understanding the origins of life, to becoming a space-faring civilization, will come from our minds. Being able to understand minds well enough to build them will be transformational in a way that no prior technology could be. No longer will we just be building tools that do what we want -- instead we will have harnessed the process of thought itself. Many people have written about the importance of AGI -- recommended reading includes Life 3.0 -- Being Human in the Age of Artificial Intelligence by Max Tegmark, Superintelligence by Nick Bostrom, and Homo Deus by Yuval Noah Harari. AGI will be the most important technology ever developed.



No-one really knows. Modern science is still far from understanding how our minds work. However our understanding is growing, and many in the scientific community feel that there is a good chance that we could see the first AGIs as early as 2030.

So why work on it now? The simple answer is that if you want the future to be a certain way, you can’t sit back and watch others build it. As Alan Kay said, "The best way to predict the future is to invent it." At Sanctuary we want a future of economic and social symbiosis -- a vibrant culture where humans and many sorts of sentient machines co-exist peacefully. So we have begun building that future, the best we know how.



We realize bringing a new form of life into the world - especially one that is so similar to us - raises non-trivial ethical, legal, and social questions. We are dedicating significant time and effort to tackling these issues - both internally and through consultation with external experts. Even though AGI is far from a fully realized technology today, we recognize the urgency of starting to address these issues in the early stages, so that our understanding of how to integrate AI into society in a beneficial way can mature alongside the technology.