Non-alcoholic steatohepatitis (NASH), a severe form of fatty liver disease, is projected to become the leading cause of liver transplants globally. Despite advances in diagnostics, the lack of continuous, personalized patient engagement remains a key barrier to effective prevention and care. This post explores an innovative solution: MirrorLiver-MCP, a generative AI-powered conversational digital twin integrated with modular clinical pathways (MCP) to transform liver health management. As a researcher and AI practitioner, I've spent the past few years exploring how generative models, those that create text, dialogue, or even medical hypotheses, can move beyond novelty and become embedded in clinical workflows. The idea behind MirrorLiver-MCP arose from a simple question: "What if every patient had an AI-powered twin that could proactively coach them through lifestyle-based care before reaching irreversible liver damage?"

The Twelfth International Conference on Learning Representations (ICLR2024) took place from 7-11 May in Vienna. The program included workshops, contributed talks, affinity group events, and socials. There were also seven invited talks that covered a broad range of topics. In this post, we give a summary of the talk by Priya Donti. Priya's research focuses on machine learning for forecasting, optimization, and control in power grids. She is an Assistant Professor and the Silverman (1968) Family Career Development Professor at MIT.

The rapid emergence of generative Language Models (LMs) has led to growing concern about the impacts that their unexamined adoption may have on the social well-being of diverse user groups. Meanwhile, LMs are increasingly being adopted in K-20 schools and one-on-one student settings with minimal investigation of potential harms associated with their deployment. Motivated in part by real-world/everyday use cases (e.g., an AI writing assistant) this paper explores the potential psychosocial harms of stories generated by five leading LMs in response to open-ended prompting. We extend findings of stereotyping harms analyzing a total of 150K 100-word stories related to student classroom interactions. Examining patterns in LM-generated character demographics and representational harms (i.e., erasure, subordination, and stereotyping) we highlight particularly egregious vignettes, illustrating the ways LM-generated outputs may influence the experiences of users with marginalized and minoritized identities, and emphasizing the need for a critical understanding of the psychosocial impacts of generative AI tools when deployed and utilized in diverse social contexts.

Priya Lakhani is an entrepreneurial adrenaline junkie. She has founded two successful companies, won numerous business awards, and acted as an advisor to the Department of Business, Innovation, and Skills. Her life's aim is to change the world for the better, and to that end she has funded meals in India, vaccinations in Africa, and her current project improves education for thousands of UK children. Remarkably, she has achieved all this despite a constant barrage of people telling her to "quit and go home". In this podcast Priya takes us through her astonishing career, from unloading boxes of sauce outside Victoria Street Station, to writing a book for preschoolers.

There are many different ways to do image recognition. Google recently released a new Tensorflow Object Detection API to give computer vision everywhere a boost. I added a second phase for this project where I used the Tensorflow Object Detection API on a custom dataset to build my own toy aeroplane detector. So what was the experience like? First lets understand the API.

In this tutorial I want to show how you can implement a skip gram model in tensorflow to generate word vectors for any text you are working with and then use tensorboard to visualize them. I found this exercise super useful to i) understand how skip gram model works and ii) get a feel for the kind of relationship these vectors are capturing about your text before you use them downstream in CNNs or RNNs. I trained a skip gram model on text8 dataset which is collection of English Wikipedia articles. I used Tensorboard to visualize the embeddings. Tensorboard allows you to see the whole word cloud by using PCA to select 3 main axis to project the data.

Path planning is the brain of a self driving car. It is the module that tries to replicate the thinking and decision making we humans do while driving -- read the map, analyze our environment (other vehicles and pedestrians) and decide the optimal action based on safety, speed and traffic rules. For my first project of term 3 in the Udacity Self Driving Car Engineer Nanodegree program, I implemented a highway path planner to drive a car on a simulated high way with multiple lanes and traffic. See the gif and video below. The simulator provides us with current position and velocity of our car and traffic at each time step back and in turn we send to the simulator the next x and y positions so it can drive the car there.

The question for 2018 is not "Why Artificial Intelligence?" but "How Artificial Intelligence?". In every major industry, AI is approaching a tipping point of widespread adoption. Gartner placed machine learning at the top of the Hype Cycle in late 2016. Since then, organizations everywhere have begun to sort through generalized excitement over this new technology as they seek to implement it. For example, consider that within the Feedzai's industry - AI for banking fraud - 65% of financial institutions told Aite in 2017 that their priority for investment in machine learning for fraud was "very high."

In this tutorial I want to show how you can implement a skip gram model in tensorflow to generate word vectors for any text you are working with and then use tensorboard to visualize them. I found this exercise super useful to i) understand how skip gram model works and ii) get a feel for the kind of relationship these vectors are capturing about your text before you use them downstream in CNNs or RNNs. I trained a skip gram model on text8 dataset which is collection of English Wikipedia articles. I used Tensorboard to visualize the embeddings. Tensorboard allows you to see the whole word cloud by using PCA to select 3 main axis to project the data.