Good afternoon from the super-polluted East Coast of the U.S.
Fires in Canada that have been burning for days have made the air unhealthy. This is a stark reminder to everyone (if one was ever needed) that borders can’t be enforced when it comes to weather patterns, pollution, and climate change.
I’m not going to post a photo of Washington Monument covered in orangish smog. Pollution is a serious matter, but in a few days, the air in Washington D.C. will be back to normal and the skies will be blue again. Cities like Beijing, New Delhi, and Lima are not so fortunate. People live in and breathe toxic fumes day in and day out.
People are wearing masks outdoors. The same ones that some of them wore indoors when COVID-19 was a a bigger threat than it is today (at least to vaccinated people). With COVID-19 we could not see the threat and we wore masks to protect others. In this case, it is pretty obvious that the haze is toxic: we’re trying to protect ourselves.
Some people are not venturing outside at all. Others are spending time in idling cars with the air-conditioning on instead of walking.
How much fake science will AI generate?
This week I wrote about the worry that generative AI will be used to create fake science. You can read the entire column here, but here are the main points.
AI-generated content can lead to fake science - including the inclusion of fabricated references and data, undermining research accuracy. The potential misuse of generative AI to introduce fictitious work and produce fake scientific papers is concerning.
Misuse of generative AI tools, like ChatGPT, is already evident in professions relying on specialized knowledge, such as law.
Peer-review, a cornerstone of the scientific process, relies on the assumption of authors' honesty and reviewers' expertise. The rise of paper mills has made it increasingly challenging to detect fraudulent scientific papers, and generative AI adds to the problem.
It's only a matter of time before an AI-generated fictitious article gets published in a scientific journal.
Generative AI can be used to craft scientific papers, but guidelines are needed to ensure transparency and accountability.
AI systems may aid in detecting scientific fraud, but the human expertise and judgment of reviewers and editors remain essential to getting the science right.
Will a liar take your job?
I appeared on “Connected” (a national news channel in Taiwan) to talk about generative AI and what it means for all of us. It was past midnight and I was bit under the weather, but do watch and let me know if I sound coherent enough to pass off as a chatbot or I’m just human. ;-)
Is the National Park in India too small for cheetahs?
This story in Nature cite scientists who certainly seem to think so.
“Nothing in biology makes sense except in the light of evolution.” observed Nobel Laureate Theodosius Dobzhansky.
Probably the same could be said of the unifying effect of the periodic table and chemistry.
Now, Nature reports that the periodic table and evolution will not be included in the curriculum for students under the age of 16 in some schools in India.
Final thoughts
Like many of you I was deeply saddened by the tragedy that happened last week. I wrote this post on LinkedIn (which resonated with many).
A high-speed passenger train crashed into a freight train near the part of India where I grew up. To add to the tragedy, a third train barreled full speed into the other two trains. A friend of my wife was on one of the trains when the accident occured. Fortunately, she and her family are unhurt, though understandably traumatized from witnessing death and destruction. Hundreds of other people were not so lucky.
Trains are not a new technology. They do not fly in the air. There are not hundreds of trains with different drivers that cross-cross multi-lane freeways. Passenger trains are not dropped on Mars and controlled from Earth with long lags.
Technology is more than self-driving cars, large language models, and hyperscale datacenters. The real world is chaotic and complex with decaying physical infrastructure, unequal access to limited resources, climate change, and people needing to make split-second decisions for which they are often not adequately equipped.
I'm an optimist by nature. We don't know all the details of why three trains crashed killing so many people but I hope this tragedy will serve as a wake-up call.
Take care,
Anirban
Thanks Anirban!!
I wanted to thank you for all your substack articles.