Independent Researcher (Remote)
San Francisco, CA /
At Generally Intelligent, our remote team is highly independent. This role is about conducting research independently into subjects that are of particular interest to us. Your sole responsibility will be to advance the frontier of human knowledge in a particular area that you are excited about.
Please note: this is an extremely challenging role for which we will hire quite selectively. This role requires both significant prior research experience, as well as comfort and familiarity with working remotely. It also requires a significant overlap between the subjects in which you are interested and the types of questions that we currently want to investigate.
Example research questions
• What are large transformers fundamentally doing? They clearly pull out some very interesting patterns about the world. How can we describe these patterns? Could we create these patterns much more efficiently?
• How can weakly / latent causal factors of the world be discovered in a mostly self-supervised fashion? How can we create models that find information that is more closely correlated with the underlying generative models of the world? How can that information help make agents that learn more quickly and are more robust to distributional shifts?
• How can we create smaller, more efficient versions of very large models? Using retrieval, hard visual attention, and other similar techniques, can we create models with effectively the same performance, yet at ≤ 1% of the computational cost?
• How can we conduct network architecture search in a practical, feasible sense, without having to define some super restricted space of architectures? What is the right formulation of the search space of architectures?
• How can we make extremely low quality versions of many of the networks that exist today (ex: Nerf, 3D mapping, image generation, etc)? Humans are clearly incapable of the same level of detail that many of these networks achieve, yet we also clearly understand some poor approximation of many of them--how are we able to do so many things, so poorly?
• Your question here! Don’t feel the need to be anchored to one of our above questions.
• Driven, self-motivated, and passionate about your research. We want people who are driven to answer scientific questions.
• An effective, independent researcher. Because the role is remote, you should be comfortable working relatively alone and independently.
• A very capable software engineer. You will need to implement and run your own experiments, so you should be very comfortable writing complex, bug-free machine learning code.
• A good communicator. You will need to communicate your results and questions to both the rest of the team, and the broader world via papers and blog posts, both of which require and ability to explain complex ideas in a succinct, easy-to-understand manner.
• Work directly on creating software with human-like intelligence.
• Generous compensation, equity, and benefits.
• $20K+ yearly budget for self-improvement: coaching, courses, conferences, etc.
• Actively co-create and participate in a positive, intentional team culture.
• Spend time learning, reading papers, and deeply understanding prior work.
How to apply
All submissions are reviewed by a person, so we encourage you to include notes on why you're interested in working with us. If you have any other work that you can showcase (open source code, side projects, etc.), certainly include it! We know that talent comes from many backgrounds, and we aim to build a team with diverse skillsets that spike strongly in different areas.
We try to reply either way within a week or two at most (usually much sooner).
We started Generally Intelligent because we believe that software with human-level intelligence will have a transformative impact on the world. We’re dedicated to ensuring that that impact is a positive one.
We have enough funding to freely pursue our research goals over the next decade, and our backers include Y Combinator, researchers from OpenAI, Astera Institute, and a number of private individuals who care about effective altruism and scientific research.
Our research is focused primarily on self-supervised and generative video and audio models. We’re excited about opportunities to use simulated data, network architecture search, and good theoretical understanding of deep learning to make progress on these problems. We take a focused, engineering-driven approach to research.