Thao Trang Nguyen

Exciting things I learned and read during the week (14 Nov – 20 Nov):

1. We’re getting a better idea of AI’s true carbon footprint

“Large language models (LLMs) have a dirty secret: they require vast amounts of energy to train and run”

“Hugging Face estimated that BLOOM’s training led to 25 metric tons of carbon dioxide emissions. But, the researchers found, that figure doubled when they took into account the emissions produced by the manufacturing of the computer equipment used for training, the broader computing infrastructure, and the energy required to actually run BLOOM once it was trained. “

“While that may seem like a lot for one model—50 metric tons of carbon dioxide emissions is the equivalent of around 60 flights between London and New York—it’s significantly less than the emissions associated with other LLMs of the same size. This is because BLOOM was trained on a French supercomputer that is mostly powered by nuclear energy, which doesn’t produce carbon dioxide emissions. Models trained in China, Australia, or some parts of the US, which have energy grids that rely more on fossil fuels, are likely to be more polluting.”

“It might also encourage people to shift toward more efficient ways of doing AI research, such as fine-tuning existing models instead of pushing for models that are even bigger, says Luccioni. “

2. How to Read the Tree Leaves

” The textbook used in that course, Brian Capon’s “Botany for Gardeners: An Introduction to the Science of Plants,” has sold more than 260,000 copies since it was published in 1990. “

Read More

Exciting things I learned and read during the week (10 Oct – 15 Oct):

1. Wonking Out: What’s Really Happening to Inflation?

“The New York Fed has an index, the Underlying Inflation Gauge, that is supposed to provide a better gauge of underlying inflation, and it looks a little better than the standard numbers:”

“But I’m not sure how well this gauge deals with my concerns. These days I often look at wages; here’s the three-month change in average wages, measured at an annual rate:”

“This measure suggests some cooling off, and it’s only running a bit more than one percentage point above prepandemic levels. So it paints a different picture than the consumer price numbers.

So what’s the moral of the story? Basically, simple rules for assessing where inflation is right now are broken. We’re in judgment territory — and that leaves lots of room for argument.”

2. Asia Sails Into Headwinds From Rate Hikes, War, and China Slowdown

“Waning momentum reflects three formidable headwinds, which may prove to be persistent:

• A sharp tightening of financial conditions, which is raising government borrowing costs and is likely to become even more constricting, as central banks in major advanced economies continue to raise interest rates to tame the fastest inflation in decades. Rapidly depreciating currencies could further complicate policy challenges.
  
• Russia’s invasion of Ukraine, which is still raging and continues to trigger a sharp slowdown of economic activity in Europe that will further reduce external demand for Asian exports.
  
• China’s strict zero-COVID policy and the related lockdowns, which, coupled with a deepening turmoil in the real estate sector, has led to an uncharacteristic and sharp slowdown in growth, that in turn is weakening momentum in connected economies.”

“In Vietnam, which is benefitting from its growing importance in global supply chains, we expect 7 percent growth and a slight moderation next year. The Philippines is forecast to see a 6.5 percent expansion this year, while growth will top 5 percent in Indonesia and Malaysia.”

3. Latin America Faces a Third Shock as Global Financial Conditions Tighten

4. In the Battle With Robots, Human Workers Are Winning

“First, humans have been underestimated. It turns out that we (well, many of us) are really amazing at what we do, and for the foreseeable future we are likely to prove indispensable across a range of industries, especially column-writing. Computers, meanwhile, have been overestimated. “

“It was a recent paper by Michael Handel, a sociologist at the Bureau of Labor Statistics, that helped me clarify the picture.

His upshot: Humans are pretty handily winning the job market. Job categories that a few years ago were said to be doomed by A.I. are doing just fine. The data show “little support” for “the idea of a general acceleration of job loss or a structural break with trends pre-dating the A.I. revolution,” Handel writes.”

“How did radiologists survive the A.I. invasion? “

“One is that humans still routinely outperform machines — even if computers can get very good at spotting certain kind of diseases, they may lack data to diagnose rarer conditions that human experts with experience can easily spot. “

“Langlotz concluded that “Will A.I. replace radiologists?” is “the wrong question.” Instead, he wrote, “The right answer is: Radiologists who use A.I. will replace radiologists who don’t.”

5. The economics of Costco rotisserie chicken

6. Are rats with human brain cells still just rats?

“A few months after they’d been implanted, the human cells made up around a sixth of the rats’ brains and appeared to have a role in controlling the animals’ behavior. Which invites the question: Are these animals still 100% rat?”

“It’s a tricky one. The scientists behind the work argue that there’s nothing really human about these rats. Throughout the study, the team examined the rats to see if those with human cells were any smarter, or experienced more suffering, than rats that didn’t receive organoid transplants. They found no sign of human traits or behaviors.”

“It can be difficult to pinpoint exactly what it is about us that makes us special, but the consensus is that it has something to do with our brains, which are larger and more complex than those of other animals. It is our brains that allow us to think, feel, dream, rationalize, form social bonds, plan our futures, and, more generally, experience consciousness and self-awareness. Could rodents with human brain cells have these same experiences?”

After being away for a while, now Interesting stuffs is back. Exciting things I learned and read during the week (12 Sept – 16 Sept):

1. We Need to Talk About How Good A.I. Is Getting

“What’s impressive about DALL-E 2 isn’t just the art it generates. It’s how it generates art. These aren’t composites made out of existing internet images — they’re wholly new creations made through a complex A.I. process known as “diffusion,” which starts with a random series of pixels and refines it repeatedly until it matches a given text description. And it’s improving quickly — DALL-E 2’s images are four times as detailed as the images generated by the original DALL-E, which was introduced only last year.”

Read More

Exciting things I learned and read during the week (12 Feb – 18 Feb) beside my current hard workload of PhD:

1. Artificial intelligence challenges what it means to be creative

“When British artist Harold Cohen met his first computer in 1968, he wondered if the machine might help solve a mystery that had long puzzled him: How can we look at a drawing, a few little scribbles, and see a face? Five years later, he devised a robotic artist called AARON to explore this idea. He equipped it with basic rules for painting and for how body parts are represented in portraiture — and then set it loose making art.”

Read More