What's New In ML #7
Robot food delivery, AI humanoids playing football, detecting Covid from voices, using AI to fight tuberculosis, generating audio, startup funding
Hello friends,
Welcome to the 7th edition of What’s New In ML on Infinite Curiosity. The goal of this segment is to help you stay updated on what’s happening in the world of ML, AI, and Data. We’ll talk about real-world applications, cutting edge research, startups, and everything else that’s moving this field forward.
If you have a question, submit it here and I’ll get back to you with my thoughts. Subscribe to this newsletter to receive it in your inbox every week:
I wanted to see what a colorful duck with glasses would look like reading a newspaper sitting on a chair. So I asked DALL-E to generate it for me. Looks pretty cool.
Let’s dive in.
APPLICATIONS AND RESEARCH
Uber Eats to do robot food delivery in California and Texas
“Starting this fall, Uber and Nuro will deploy autonomous delivery vehicles in two cities: Mountain View, California, and Houston, Texas. Neither company would disclose the number of vehicles nor the expected number of customers who will participate in these early tests, but they did say they eventually hope to expand the service area to the greater Bay Area in California. The permit, issued by the California DMV, only allows the company to operate its delivery service in parts of Santa Clara and San Mateo counties, which would mean most of the Silicon Valley and its tech workers would be within its domain but not San Francisco or Oakland. That means the company will need to obtain additional permission from the DMV before expanding its service area.”
Great development for the autonomous vehicles market. It would be interesting to see how it plays out. Read more
DeepMind is teaching AI humanoids to play football
"An AI has to re-create everything human players do—even the things we don’t have to consciously think about, like precisely how to move each limb and muscle in order to connect with a moving ball—making hundreds of decisions a second."
Teaching a humanoid robot to play football (they call it soccer in the article) is not a trivial task. Read more
AI model that can detect Covid-19 in people's voices
"The researchers used a voice analysis technique called Mel-spectrogram analysis, which identifies different voice features such as loudness, variation, and power over time. The AI model used in this research claims to be more accurate than the rapid antigen tests as well as quicker and easier to use. The model aims to assist in the detection of infection in low-income countries, where PCR tests are often expensive and difficult to distribute."
Researchers from Maastricht University claim to have built a system that can detect Covid-19 from people's voices. There's lot of actual field testing that needs to be done. But this could be helpful it works as promised in the real world. Read more
Google debuts a new AI tool in the global fight against tuberculosis
"Tuberculosis kills 1.4 million people every year, primarily in places where poverty and deprivation conspire to make people uniquely vulnerable, and unable to get lifesaving care in time. Google is now joining a global fight to snuff out the disease, using AI to automate its detection — and expedite treatment — in communities where physicians are in short supply. A new study published Tuesday in Radiology, the journal of the Radiological Society of North America, found that its AI model performed as well as radiologists at detecting tuberculosis on chest X-rays."
AI is increasingly being used in healthcare. And the use cases are becoming more tangible. This is yet another example. Read more
Google introduces AudioLM, a framework for high-quality audio generation
"AudioLM maps the input audio to a sequence of discrete tokens and casts audio generation as a language modeling task in this representation space. We show how existing audio tokenizers provide different trade-offs between reconstruction quality and long-term structure, and we propose a hybrid tokenization scheme to achieve both objectives."
They're using language modeling techniques to generate audio. They say that this system can generate plausible speech continuations while maintaining speaker identity. They've also extended it to musical instruments like piano. Read more
Quantifying GitHub Copilot’s impact on developer productivity and happiness
"Developers reported that GitHub Copilot helped them stay in the flow (73%) and preserve mental effort during repetitive tasks (87%). That’s developer happiness right there, since we know from previous research that context switches and interruptions can ruin a developer’s day, and that certain types of work are draining."
GitHub Copilot helps generate code based on the developer's needs and style. It's an amazing application of AI. And it looks like they're making efforts to quantify exactly how much impact it's having on developers. Read more
AI system that makes generative models more creative
"To generate more complex images with better understanding, scientists from MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL) structured the typical model from a different angle: they added a series of models together, where they all cooperate to generate desired images capturing multiple different aspects as requested by the input text or labels. To create an image with two components, say, described by two sentences of description, each model would tackle a particular component of the image."
DALL-E can generate fantastical images based on text input. So MIT researchers developed a system to make DALL-E more creative. They've used multiple models to create more complex images with better understanding. Read more
PyTorch is moving to an independent foundation
"To accelerate progress in AI, PyTorch is moving to a new, independent PyTorch Foundation, under the Linux Foundation umbrella. The project will join the Linux Foundation with a diverse governing board composed of representatives from AMD, Amazon Web Services, Google Cloud, Meta, Microsoft Azure, and Nvidia, with the intention to expand over time. The PyTorch Foundation will act as a responsible steward for the technology and support PyTorch through conferences, training courses, and other initiatives. The foundation’s mission is to drive adoption of AI tooling by fostering and sustaining an ecosystem of open source, vendor-neutral projects with PyTorch. It will democratize state-of-the-art tools, libraries, and other components to make these innovations accessible to everyone."
This is a big development. PyTorch is one of the world's most popular ML libraries. But it was under Meta, which made many people nervous because you never know when they will pull it off the market. But with this move, it has moved to an independent foundation. It's open source and no single company can control what happens to it. Good for the ML developer community overall. Read more
STARTUP FUNDING
Awiros raises $7M Series A led by Inflexor and Exfinity for its video AI solutions
Spice AI raises $13.5M Seed led by Madrona for its platform that enables developers to build data driven applications for web3
Alcatraz AI raises $25M Series A led by Almaz for its autonomous access control solution
Diveplane raises $25M Series A led by Shield Capital for its AI-powered business solutions
Hebbia raises $30M Series A round led by Index Ventures for its document search tool
Arize AI raises $38M Series B led by TCV for its ML observability platform
PolyAI raises $40M Series B led by Georgian for its conversational AI
PathAI raises $165M Series C co-led by D1 Capital and Kaiser Permanente for its diagnostic and drug development software
If you are getting value from this newsletter, consider subscribing and sharing with your friends: