Yesterday, Microsoft's Build 2020 developer conference kicked off (remotely), and we saw the first results of Microsoft's billion-dollar investment in OpenAI, a company co-founded by Elon Musk. Microsoft announced it has developed an Azure-hosted supercomputer built expressly for testing OpenAI's large-scale artificial intelligence models. While we've seen many AI implementations focused on single tasks, like recognizing specific objects in images or translating languages, a new wave of research focuses on massive models that can perform multiple tasks at once. As Microsoft notes, that can include moderating game streams or potentially generating code after exploring GitHub. Realistically, these large-scale models can actually make AI a lot more useful for consumers and developers alike.

At its Build developer conference, Microsoft today announced that it has teamed up with OpenAI, the startup trying to build a general artificial intelligence, with -- among other things -- a $1 billion investment from Microsoft, to create one of the world's fastest supercomputers on top of Azure's infrastructure. Microsoft says that the 285,000-core machine would have ranked in the top five of the TOP500 supercomputer rankings. Because Microsoft doesn't actually tell us much more than that, except for a few more specs that say it had 10,000 GPUs and 400 gigabits per second of network connectivity per server, we'll just have to take Microsoft's and OpenAI's word for this. To be in the top five of supercomputers, a machine would currently have to reach more than 23,000 teraflops per second. It's also worth noting that the No. 1 machine, the IBM Power System-based Summit, reaches over 148,000 teraflops, so there is quite a wide margin here.

Last year, Microsoft invested $1 billion in Open AI, a non-profit co-founded by Elon Musk that focuses on the development of human-friendly artificial intelligence. Microsoft announced that it has developed an Azure-hosted supercomputer built expressly for testing OpenAI's large-scale artificial intelligence models. While we've seen many AI implementations focused on single tasks, like recognizing specific objects in images or translating languages, a new wave of research is focused on massive models that can perform multiple tasks at once. As Microsoft notes, that can include moderating game streams or potentially generating code after exploring GitHub. Realistically, these large-scale models can actually make AI a lot more useful for consumers and developers alike.

Over the last decade, there has been significant progress in the field of machine learning for de novo drug design, particularly in deep generative models. However, current generative approaches exhibit a significant challenge as they do not ensure that the proposed molecular structures can be feasibly synthesized nor do they provide the synthesis routes of the proposed small molecules, thereby seriously limiting their practical applicability. In this work, we propose a novel forward synthesis framework powered by reinforcement learning (RL) for de novo drug design, Policy Gradient for Forward Synthesis (PGFS), that addresses this challenge by embedding the concept of synthetic accessibility directly into the de novo drug design system. In this setup, the agent learns to navigate through the immense synthetically accessible chemical space by subjecting commercially available small molecule building blocks to valid chemical reactions at every time step of the iterative virtual multi-step synthesis process. The proposed environment for drug discovery provides a highly challenging test-bed for RL algorithms owing to the large state space and high-dimensional continuous action space with hierarchical actions. PGFS achieves state-of-the-art performance in generating structures with high QED and penalized clogP. Moreover, we validate PGFS in an in-silico proof-of-concept associated with three HIV targets. Finally, we describe how the end-to-end training conceptualized in this study represents an important paradigm in radically expanding the synthesizable chemical space and automating the drug discovery process.

Eight years ago a machine learning algorithm learned to identify a cat--and it stunned the world. A few years later AI could accurately translate languages and take down world champion Go players. Now, machine learning has begun to excel at complex multiplayer video games like Starcraft and Dota 2 and subtle games like poker. AI, it would appear, is improving fast. But how fast is fast, and what's driving the pace?

The future is now people. Not only do we have pandemic-proof rave suits being designed, we also now might be on the precipice of having music released made with Artificial Intelligence thanks to the latest development from OpenAI. Aptly titled'Jukebox', the new model is now able to generate genre-specific music. According to OpenAI's website, Jukebox is "a neural net that generates music, including rudimentary singing, as raw audio in a variety of genres and artist styles." Using over 1.6million songs as their dataset, Jukebox is able to use a song provided as input, and generate a sample produced from scratch in specific genres as output.

Microsoft's AI ethics committee helped craft internal Department of Defense contract policy, and G20 member nations wouldn't have passed AI ethics principles if it weren't for Japanese leadership. Published Tuesday, the UC Berkeley Center for Long-Term Cybersecurity (CLTC) case study examines how organizations are putting AI ethics principles into practice. Ethics principles are often vaguely phrased rules that can be challenging to translate into the daily practices of an engineer or other frontline worker. CLTC research fellow Jessica Cussins Newman told VentureBeat that many AI ethics and governance debates have focused more on what is needed, but less on the practices and policies necessary to implement goals enshrined in principles. The study focuses on OpenAI's rollout of GPT-2; the adoption of AI principles by OECD and G20; and the creation of the AI, Ethics, and Effects in Engineering and Research (AETHER) committee at Microsoft.

Artists may need to start competing with -- or embracing -- computer-made songs and soundtracks in the near future, if a new AI music generator shows any indication of what could come next for the music industry. Researchers at artificial intelligence lab OpenAI have released Jukebox, an open-source algorithm that can generate music, complete with lyrics, vocals, and a soundtrack. All the algorithm needs is a genre, an artist, and a snippet of lyrics, and Jukebox can create song samples that can be realistic and quite catchy. OpenAI's music generator runs on the same sort of machine-learning technology used to create deepfakes and employed by the slew of sites that popped up in 2019 generating fake memes, fake Airbnb listings, and fake cats. Jukebox produces its AI creations using artificial neural networks that train a computer to learn from an influx of data.

OpenAI today announced it will begin tracking machine learning models that achieve state-of-the-art efficiency, an effort it believes will help identify candidates for scaling and achieving top overall performance. To kick-start things, the firm published an analysis suggesting that since 2012, the amount of compute needed to train an AI model to the same performance on classifying images in a popular benchmark -- ImageNet -- has been decreasing by a factor of 2 every 16 months. Beyond spotlighting top-performing AI models, OpenAI says that publicly measuring efficiency -- which here refers to reducing the compute needed to train a model to perform a specific capability -- will paint a quantitative picture of algorithmic progress. It's OpenAI's assertion that this in turn will inform policy making by renewing the focus on AI's technical attributes and societal impact. "Algorithmic improvement is a key factor driving the advance of AI. It's important to search for measures that shed light on overall algorithmic progress, even though it's harder than measuring such trends in compute," OpenAI wrote in a blog post.

OpenAI recently launched Jukebox, a model that generates music with singing in the raw audio domain. As a generative model for music, Jukebox can handle the long context of raw audio using an autoencoder. Jukebox's autoencoder processes the audio files using a multiscale VQ-VAE to compress it to discrete codes and modeling those using autoregressive Transformers. Provided with a genre, artist, and lyrics as input, Jukebox can output a new music sample produced from scratch. This is a type of innovation that expands the boundaries of generative models to a new level.