Ilya Sutskever NeurIPS Keynote Ilya Sutskever's NeurIPS keynote is up - an interesting reflection on the decade since the seminal “Sequence to Sequence” work that helped ignite the modern era of large-scale neural NLP: This is a comprehensive session where Ilya recounts how the original Seq2Seq approach established a template for present-day AI: big models plus big data equals breakthroughs. Over time, this scaling principle was validated far beyond translation, culminating in today’s GPT-style models. However, it seems we are exhausting the “fossil fuel” of internet-scale data, and future progress will hinge on new techniques - eg. agents interacting with their environments, generating synthetic data, and improved reasoning capabilities. |
|
|
|---|
|
Google Releasing Gemini 2.0 Google drops the mic last week releasing Gemini 2.0 with a bunch of new features on their AI studio, doubling down towards "agentic" AI with multimodal input/output: As expected Google is building on the initial Gemini 1.x foundation, extending long-context and multimodality capabilities, improves latency and performance, and introducing features like native image and audio generation. Something that comes across as novel is the integration into Google’s products and ecosystem, however the race continues to move at breakneck speed so we can only expect similar pace from the tech ecosystem. |
|
|
|---|
|
Tiktok's Recommendation System Tiktok has published the architecture behind their recommendation system called "Monolith", a real-time massive-scale recommendation system designed specifically to address production challenges such as large-scale, sparse, and dynamic feature spaces: The paper provides interesting insights such as collisionless embedding table based on Cuckoo hashing, which enables dynamic inclusion and eviction of new model features. Monolith tightly integrates training and serving which they highlight as one of the reasons they can allow for fast online updates so that the model can adapt to changing user behavior within minutes. |
|
|---|
|
META FAIR Major Releases Meta Fundamental AI Research (FAIR) has quietly shared a huge release last week with new open-source AI systems across CLIP, Motivo and Seal: Meta released Motivo, a foundation model that enables embodied humanoid agents to efficiently solve complex tasks without additional training. Meta also released Video Seal, a robust watermarking solution for videos that remains intact through common transformations. They’ve also introduced Flow Matching, a hierarchical byte-level tokenizer-free approach (Dynamic Byte Latent Transformer) for generative modeling. Additionally, they released a new version of Meta CLIP, which improves on previous versions for vision-language alignment. Quite surprising and exciting to see so much movement from META's research arm furthering the research ecosystem across quite a few of these interesting areas. |
|
|
|---|
|
OpenAI Releases Text-to-Video OpenAI has finally released their Text-to-Video SORA model as a public offering! As per the usual naming convention, this comes with Sora Turbo, focusing on fast generation of higher fidelity videos across multiple aspect ratios, and up to 20-second. These services continue to surprise us with the quality of the video generation, certainly still with quite some limitations (such as many posts showing the limits when rendering scenes from gymnastics, etc). The model is still imperfect, struggling with complex sequences and realistic physics - however it is great to see finally OpenAI is releasing to encourage community input and iteration norm-setting, and responsible use. |
|
|
|---|
|
Upcoming MLOps Events The MLOps ecosystem continues to grow at break-neck speeds, making it ever harder for us as practitioners to stay up to date with relevant developments. A fantsatic way to keep on-top of relevant resources is through the great community and events that the MLOps and Production ML ecosystem offers. This is the reason why we have started curating a list of upcoming events in the space, which are outlined below. Upcoming conferences where we're speaking:
Other upcoming MLOps conferences in 2024:
In case you missed our talks:
|
|
|---|
| | |
Check out the fast-growing ecosystem of production ML tools & frameworks at the github repository which has reached over 10,000 ⭐ github stars. We are currently looking for more libraries to add - if you know of any that are not listed, please let us know or feel free to add a PR. Four featured libraries in the GPU acceleration space are outlined below. - Kompute - Blazing fast, lightweight and mobile phone-enabled GPU compute framework optimized for advanced data processing usecases.
- CuPy - An implementation of NumPy-compatible multi-dimensional array on CUDA. CuPy consists of the core multi-dimensional array class, cupy.ndarray, and many functions on it.
- Jax - Composable transformations of Python+NumPy programs: differentiate, vectorize, JIT to GPU/TPU, and more
- CuDF - Built based on the Apache Arrow columnar memory format, cuDF is a GPU DataFrame library for loading, joining, aggregating, filtering, and otherwise manipulating data.
If you know of any open source and open community events that are not listed do give us a heads up so we can add them! |
|
|---|
| | |
As AI systems become more prevalent in society, we face bigger and tougher societal challenges. We have seen a large number of resources that aim to takle these challenges in the form of AI Guidelines, Principles, Ethics Frameworks, etc, however there are so many resources it is hard to navigate. Because of this we started an Open Source initiative that aims to map the ecosystem to make it simpler to navigate. You can find multiple principles in the repo - some examples include the following: - MLSecOps Top 10 Vulnerabilities - This is an initiative that aims to further the field of machine learning security by identifying the top 10 most common vulnerabiliites in the machine learning lifecycle as well as best practices.
- AI & Machine Learning 8 principles for Responsible ML - The Institute for Ethical AI & Machine Learning has put together 8 principles for responsible machine learning that are to be adopted by individuals and delivery teams designing, building and operating machine learning systems.
- An Evaluation of Guidelines - The Ethics of Ethics; A research paper that analyses multiple Ethics principles.
- ACM's Code of Ethics and Professional Conduct - This is the code of ethics that has been put together in 1992 by the Association for Computer Machinery and updated in 2018.
If you know of any guidelines that are not in the "Awesome AI Guidelines" list, please do give us a heads up or feel free to add a pull request!
|
|
|---|
| | |
| | | | The Institute for Ethical AI & Machine Learning is a European research centre that carries out world-class research into responsible machine learning. | | | | |
|
|
|---|
|
|
This email was sent to You received this email because you are registered with The Institute for Ethical AI & Machine Learning's newsletter "The Machine Learning Engineer"
|
| | | | |
|
|
|---|
|
© 2023 The Institute for Ethical AI & Machine Learning |
|
|---|
|
|
|
The Institute for Ethical AI & Machine Learning
