Sign in Subscribe

Topic

Artificial Intelligence

A collection of 64 issues

What has not changed? The case on Nvidia

It was impossible to open the news before Nvidia's recent earnings and not read some speculations regarding Nvidia earnings. The media portrayed Nvidia earnings as THE indicator for where AI is going. Articles regarding Nvidia earnings read like suspense movies. From the market open on February 20 to

Joyce’s picks: musings and readings in AI/ML, February 19, 2024

📌 Musings and Readings "CDAO Launches First DOD AI Bias Bounty Focused on Unknown Risks in LLMs" (U.S. Department of Defense) "Video generation models as world simulators" (OpenAI) "Diagnostic reasoning prompts reveal the potential for large language model interpretability in medicine" (Nature) 👏 Deals Rogo,

Joyce’s picks: musings and readings in AI/ML, February 12, 2024

📌 Musings and Readings “How symmetry can come to the aid of machine learning” (MIT) “Bard becomes Gemini” (Google) “Acoustic Model Fusion for End-to-End Speech Recognition” (Apple) “Shaping the future of advanced robotics” (Deepmind) “Building geospatial foundation models via continual pretraining” (Amazon) 👏 Deals Nauticus Robotics, a Houston-based company developing subsea robots

Multimodal Learning

Multimodal learning is an approach in machine learning that involves the integration and processing of information from multiple modalities or sources. A modality refers to a distinct type of data, such as text, images, audio, video, or any other form of sensory input. In multimodal learning, the goal is to

Joyce’s picks: musings and readings in AI/ML, February 5, 2024

📌 Readings Amazon earnings press release with strategic initiative details (link) Meta earnings call transcript (link) Microsoft product announcement (link) Alphabet earnings call transcript (link) AMD earnings slides (link) 👏 Deals Carpl, a CA-based radiology AI tech startup, has raised $6m in financing. Cohere Health, a Boston-based startup which provides intelligent prior-authorization

Google's Gemini

Google launched Gemini today. Gemini 1.0 is a multimodal model that has been trained on image, audio, video and text data and is intended to be generalizable across different data modalities (e.g., you give an image, it can return text-based description of the image). Gemini comes in three

OpenAI Board Should Resign

What happened at OpenAI was crazy on many levels. While I am an outsider to OpenAI's inner workings, I will go out on a limb and say that the people who made the decision to oust Sam Altman probably do not have the right experience for making this

New Industrial Age

Where we are going in Generative AI will fundamentally change how knowledge work will be done, full stop. So we must ask ourselves, what is the future of knowledge work and knowledge workers? Customer services and marketing have already been disrupted. Research, innovation service, creative writing, paralegal, early career roles

Joyce’s picks: readings in ML/AI, October 30, 2023

“Sweeping new Biden order aims to alter the AI landscape” (Politico) “G7 to agree AI code of conduct for companies” (Reuters) “How AI can help to save endangered species” (Nature) “A More Reliable Wikipedia Could Come from AI Research Assistants” (Scientific American) “Take a Step Back: Evoking Reasoning via Abstraction

Evaluating Gen AI Applications - Part I

When I worked in mergers & acquisitions way back, one of the things to do was generating project code names to help preserve the confidentiality of these highly sensitive projects. I remember that we have gone through all kinds of names, drawing inspiration from mountains, park names, bird names, car