Why we need to do a better job of measuring AI’s carbon footprint
Asking tech companies to provide more data on the climate impacts of building, training, and using AI is a start. We should also avoid obsessing over building larger and larger AI models and try to figure out ways to do AI research with more energy-efficient methods, such as fine-tuning the models. existing image.
Inside Alphabet X’s new effort to fight climate change with AI and seagrass
The MIT Technology Review got a sneak peek at Tidal, a new climate change mitigation project by X, the hypercar division of Google’s parent company, Alphabet. Tidal uses cameras, computer vision and machine learning to track how much carbon is stored in the biomass of the oceans. This is part of an effort to improve our understanding of aquatic ecosystems to inform and encourage efforts to protect the oceans amid growing threats from pollution. , overfishing, ocean acidification and global warming.
With projects like Tidal, X is creating tools to ensure that industries can do more. to address environmental hazards and ecosystems that may exist in a hotter, more inhospitable world. It also relies heavily on the parent company’s areas of strength, on Alphabet’s robotics expertise as well as its ability to glean insights from massive amounts of data using artificial intelligence. Read James Temple story about it.
Bits and bytes
Elon Musk is starting to see the consequences of laying off AI teams
When the billionaire took over Twitter, he fired half of the company’s employees, including the machine learning teams that worked to ensure that the platform’s infrastructure was safe, secure, and trusted. The ethics-AI team and those working on the infrastructure are among those who have left. The results were almost immediate: the site was slowly starting to break down. We spoke to a former Twitter engineer to see how successful it is likely to be. (MIT Technology Review)
A lawsuit could rewrite AI copyright rules
In the first class action lawsuit in the US for training AI systems, Microsoft, GitHub and OpenAI are being sued for allegedly violating copyright law by copying open source code using AI. GitHub Copilot scans web pages for code and, like large language models, retrieves what it gathers in its database without attributing the original source. A lawsuit challenging the legitimacy of this model could have a massive stimulant effect on other AI systems trained by web scraping, from text generation to image-generating AI. (precipice)
Are the US and China really in a cold war over AI?
This is a really interesting series that unpacks some of the problematic stories surrounding the AI development race between the US and China. (protocol)
Amazon’s new robot can handle most items in its inventory
The new robot, called Sparrow, can pick up items from shelves or bins to pack into boxes. This has traditionally been too complicated for robots, because there are so many different types of objects with different shapes and sizes. Amazon’s robot uses machine learning and cameras to identify objects. This can help speed up warehouse operations. (Wired)
super model generator
A new text-to-image AI called Aperture is reported about to fall This week from Lexica, and it seems to be able to produce super realistic looking supermodel photos. I’m curious to see this model in action as other popular image-generating AIs, such as DALL-E and Stable Diffusion, are having a hard time generating fingers and handsas well as human faces like no other melt in the sun.