Aptos Research Vision: Cutting-edge Research Meets World-class Engineering
Make motivation real again
Consider the life cycle of a research project at a conventional industrial lab. Quite often it proceeds in the following order:
- Look for a potentially publishable direction in the field.
- Focus on publishing a result, declaring success upon deriving the result and getting acceptance, and
- Lastly, look for a real-world application as an added bonus. For example, chat with engineers and try to convince them that the published result is so great that they should use it.
At Aptos Labs, we flip this script. Here:
- We first identify hard problems that block us from achieving the performance, security and usability in the Aptos blockchain and its surrounding ecosystem.
- We then focus on solving these problems, tracking our effort all along to drive significant production impact, and
- Publish and celebrate the achieved results at top-tier venues.
We believe in this approach because the current landscape in the field of blockchain & Web 3.0 is full of significant problems, still open and unresolved, standing in the way of widespread adoption for billions of people across a broad set of use cases. More importantly, these problems span disciplines such as cryptography, distributed systems, programming languages, game theory, databases, networking and concurrency. To investigate and prioritize these problems, we rely on our close connection with our world-class engineering team and their intimate knowledge of the needs of the Aptos blockchain. Hence, we are in a unique and privileged position where solutions to most of the hard problems that we face are most likely going to be novel and publishable.
We understand that the research process can be unpredictable and non-linear. We may not solve every problem under all its constraints. Still, we search for solutions that do not fall apart when faced with the hard reality of a practical deployment. Our approach is to “find a way, or make one”. We do not get stuck with extremes, such as stronger, unsolvable variants of the problem or oversimplified solutions that only work in convoluted academic settings.
Make publication worthy again
We aspire to advance the state of the art in blockchain technology. We welcome collaboration with others across the industry and academia who share the same goals. We dedicate significant effort to explain our results and share our knowledge via less technical and more educational mediums such as blog posts, talks and seminars.
We strive to publish results that we are proud of at the best possible venues. We never publish for the sake of publishing nor do we optimize artificial metrics.
We hold ourselves to the highest standards, requiring rigorous proofs of correctness for our protocols, and comprehensive experiments designed to leave no doubt in any of our claims. We utilize the academic review process as a tool to help us achieve this objective.
If we think our result is being unfairly judged, we start by first taking a hard look at ourselves, by asking hard questions and soliciting brutally-honest feedback from academics and other experts. We will not be blinded by appeasing the fragile ego of reviewer #2 (e.g., No, we will not reproduce our system and experiments on HyperLedger Fabric..) nor ever develop an unhealthy fixation towards a particular venue or perceived academic prestige.
In other words, if the academic review process for a top-venue acceptance becomes too political or adversarial, we conclude that the overhead might not always be worth it. Instead, we let the ultimate judge of our contributions and impact be the time itself and the objectively-measurable metrics at the massive scale of the Aptos blockchain.
Make debate open again
We believe that open and direct discussions, even passionate debates at times, are the most healthy, efficient way for making progress. Crucial to this are mutual respect and mutual agreement:
- To criticize the ideas and never the people, and
- Trust that others are following (1).
We believe that by doing so, the best ideas will emerge and the best technology will make its way to the Aptos blockchain.
Make research deploy again
We consider advancing the frontier of our knowledge as a success metric for any research project. Therefore, we are not discouraged when an approach does not lead to real-world impact. However, when an approach does pull through, we take responsibility for seeing it implemented in production.
In other words, Aptos researchers lean into engineering: coding is not beneath them, but quite the opposite.
Research scientists at Aptos Labs combine foundational knowledge of the respective fields with the ability to speak the same language as the engineers, willingness to learn and grow, and follow the desire to deliver the projects to completion, possibly landing and owning the code in production.
Research engineers are instrumental to this process. Research engineers provide the creativity and the engineering pedigree to transform novel, unconventional, and research-originated approaches into the high-performance, high-quality robust system components for the real-world deployment.