We’re excited to announce the 2023 Gradient Prize! TLDR: Submit full, finished pieces any time this calendar year — we can’t wait to see what you come up with!
To enter the contest, just fill out our writing interest form and indicate interest in the contest in your submission.
Prize allocation
TBD
Guidelines for a winning essay
The remainder of these should be considered rough, flexible guidelines for pieces that have been successful in the past. If you write something truly brilliant that contradicts these, please still feel free to submit!
- Wordcount: between 1500 and 4000 words.
- While we aren’t looking for literary genius, the piece should be well-written and express ideas clearly. The intended audience for your piece should be a technical person who is generally familiar with machine learning and big-picture issues, but not necessarily well-versed in your particular subfield.
- The topic should be relevant to the ML community as a whole -- or, make the case for why it should be relevant.
- Your piece should be grounded in solid background research or technical details, not solely opinion. We expect citations to e.g. news media and/or academic papers.
- If relevant, illustrations are encouraged, but not necessary.
- Some broad categories for potentially successful pieces:
- An in-depth introduction to X { subfield, application }
- Your position on a current debate in your (sub)field
- Make the case for X { e.g. research direction, but any opinion really }
- Examples of the sort of articles we encourage include:
- Accessible but technically informed overviews of the state of the art and its future such as The Economics of AI Today, Introduction to Artificial Life for People who Like AI, and The Promise of Hierarchical Reinforcement Learning.
- Critical and well-informed perspectives on artificial intelligence topics such as NLP's ImageNet moment has arrived, Quantifying Independently Reproducible Machine Learning, and The #BenderRule: On Naming the Languages We Study and Why It Matters.
Selection process
The process of selecting winners will be as follows.
- The core team of Gradient editors will read all submissions and compile a shortlist of finalists. Each submission will be read by at least two members of Gradient editorial staff.
- The shortlist of finalists will be anonymized and sent to our guest judge(s).
Previous Winners
The 2022 Gradient prize was one by Graph Neural Networks beyond Weisfeiler-Lehman and vanilla Message Passing, with Engaging with Disengagement and AI is Ushering In a New Scientific Revolution as runner ups. See our full announcement for more details!
The 2021 Gradient prize was won by The Imperative for Sustainable AI Systems, with Machine Translation Shifts Power and Machine Learning Won't Solve Natural Language Understanding winning the runner up prizes.
