AI, James Webb Area Telescope analysis and local weather change : NPR

ARI SHAPIRO, HOST:

It is time now for our science information roundup from our pals at NPR’s Quick Wave podcast, Regina Barber and Geoff Brumfiel. Good to have you ever right here on the finish of the yr.

REGINA BARBER, BYLINE: Hey.

GEOFF BRUMFIEL, BYLINE: Hello there.

SHAPIRO: You normally discuss us by means of three issues taking place within the science world every week, contemporary off the journals and social media. However as we speak, you are going to give us three massive issues in science from the yr 2023. I could not be extra enthusiastic about it.

BARBER: Superior. Nicely, we’ll tick by means of the highest three science threads we noticed unraveling this yr.

BRUMFIEL: We’ve all of the scientific discoveries taking place from the James Webb Area Telescope.

SHAPIRO: That is what that pile in your lap is.

BRUMFIEL: (Laughter).

BARBER: And contemporary off of the COP28 convention, we’ll discuss local weather.

BRUMFIEL: However we’ll begin with one other actually massive theme, which I am positive you are acquainted with – synthetic intelligence.

SHAPIRO: Sure. I am really not talking proper now. It’s a…

BARBER: (Laughter).

SHAPIRO: No. It isn’t an AI reproduction of my voice, however it appears like, this yr, everyone’s afraid that AI goes to come back for our jobs.

BARBER: Yep.

SHAPIRO: So what have you ever acquired for us?

BRUMFIEL: Yeah. I imply, you realize, it has been an enormous fear for lots of various fields. Visible artists have been form of spooked by these picture mills, like Lensa, Midjourney and Dall-E. Lensa’s additionally getting used to make AI social media avatars, that are, amongst different issues, getting used to push magnificence requirements to even increased, extra unobtainable ranges.

BARBER: And, in fact, the large tech house program that has many a journalist, author – you title it – in a panic was ChatGPT, the chatbot created by the corporate OpenAI. And over the course of the yr, ChatGPT acquired increasingly highly effective and higher at producing textual content. And it is not simply ChatGPT. Like, most of the high tech corporations all have their very own AI, like Microsoft’s Bing AI chatbot. That one made headlines this yr for doing every little thing from professing its love for a journalist and attempting to get him to depart his spouse to, extra just lately, linking to conspiracy theories and lies when requested about elections.

SHAPIRO: Nicely, panic apart, lots of the industries which have began to actually combine AI are within the STEM fields – science and tech – proper?

BRUMFIEL: Yeah. So let’s speak about drugs for a minute. You realize, the AI revolution was already effectively underway in drugs earlier than ChatGPT confirmed up. Folks have been growing algorithms to do issues like diagnose ailments in scans and issues like that. However with the language fashions, issues are going even additional. Some corporations are floating methods to try to streamline medical notes and affected person data. Others are rolling out packages that may generate correspondence between docs and sufferers, and that is acquired some researchers nervous. I spoke to Marzyeh Ghassemi, who’s at MIT and research AI makes use of in well being care. She cited one instance the place Microsoft simply rolled out software program to some hospitals that makes use of AI to write down messages from docs to their sufferers.

MARZYEH GHASSEMI: They’re permitting it to draft this textual content as a suggestion of how an individual ought to talk with their affected person, and that worries me.

BRUMFIEL: As a result of it actually hasn’t been examined, and we all know AI can endure from hallucinations – principally simply make stuff up. It may give deceptive info. And there is bias within the coaching units that may discriminate form of unconsciously in opposition to completely different teams of individuals. However the truth of the matter is, there’s a lot strain on docs within the medical system that these kinds of instruments are going to get tried, and that is what’s taking place.

BARBER: Yeah. And within the sciences, AI is already beginning to discover its place, particularly in fields like chemistry and biology, the place researchers have, like, an enormous variety of molecules and compounds to check. It will probably attempt to discover ones that match the researchers’ standards, and folks can synthesize the candidate chemical compounds or compounds in actual life to see whether or not they work. And a few labs are taking issues a step additional. Researchers at Lawrence Berkeley Nationwide Laboratory in California and Carnegie Mellon College in Pittsburgh have constructed robots that may do a number of the synthesis and testing themselves, doubtlessly additional chopping down on time.

SHAPIRO: And, you realize, so long as I have been alive, I’ve heard that STEM careers include assured jobs. What does it imply if AI can do superior chemistry? I imply, are scientists going to be out of labor?

BRUMFIEL: Yeah. You realize, I have been fascinated by jobs loads this yr as I have been reporting, and I maintain developing in opposition to this form of primary fact – whether or not AI is being utilized in drugs, science and even issues like surveillance and nationwide safety, it all the time looks like it comes again to the truth that it really works greatest with a human within the loop. You realize, I spoke to Sasha Luccioni, a researcher at an AI firm referred to as Hugging Face, and she or he put it this manner.

SASHA LUCCIONI: I do not see generative AI fashions changing individuals, however I can see them, you realize, serving to individuals or being utilized by individuals of their present jobs in an effort to – no matter – to go sooner or to be – you realize, to have extra inventive concepts.

BRUMFIEL: And our colleagues at Planet Cash actually noticed this firsthand. They used AI to make a sequence of episodes, and it labored. I imply, it did lots of the work for them, however it labored greatest once they have been giving it suggestions. So I believe that is one thing that – you realize, we are able to maintain out some hope that they are not simply going to take our jobs – the bots.

SHAPIRO: OK, effectively, our second massive matter is local weather. What have you ever acquired for us?

BARBER: Yeah. So 2023 was a scorching yr – so scorching that, as soon as all the information is in, it is anticipated to be introduced that this yr was the most popular on report.

SHAPIRO: And I do know scientists say, if we wish to keep away from essentially the most catastrophic results of local weather change, people have to maintain world temperatures from growing greater than 1.5 levels Celsius in comparison with preindustrial ranges. How shut are we to that quantity proper now?

BRUMFIEL: The typical temperature of the Earth during the last decade was about one diploma Celsius increased than preindustrial temperatures, so we’re undoubtedly getting shut.

BARBER: Yeah. And earlier this yr, our colleague on the local weather desk, Rebecca Hersher – she reported that to restrict warming to 1.5 levels Celsius, people must slash greenhouse fuel emissions greater than 40% by 2030.

SHAPIRO: And we’re not on observe to do this.

BRUMFIEL: Yeah. Even essentially the most bold plans to chop emissions would not get to zero by 2050. So that you’re proper – it is unlikely. However it’s not like a cliff the place people are doomed as quickly as we attain 1.5 levels Celsius. We nonetheless have lots of energy to restrict the detrimental impacts of local weather change.

SHAPIRO: Nicely, proper. A cliff implies that you simply both go over it or you do not. But when we do not hit 1.5, we’ve got to maintain it beneath two. And if we do not maintain it beneath two, we’ve got to maintain it beneath two – like, it retains getting worse, proper?

BARBER: Yeah. I imply, that’s nerve-racking. However I do wish to say it is not all doom and gloom. Like, people are taking motion across the globe. And NPR coated a few of these efforts throughout our local weather options week, like how Uruguay is utilizing wind energy and different inexperienced power sources to assist energy their grid – 98% of the nation’s grid.

SHAPIRO: OK, we’ll take hope from Uruguay. Let’s depart this planet for a second and test in on the James Webb Area Telescope, which continued sending again unbelievable photographs this yr. Gina, you have reported on what this implies for astronomy.

BARBER: Yeah. So this telescope has given astronomers a view into the early universe, like exhibiting us the earliest galaxies, black holes, stars that we have ever seen. And what they seem like is stunning to scientists as a result of they’re far more grown-up than scientists like astrophysicist Jorge Moreno anticipated.

JORGE MORENO: It is like when you went to a kindergarten and also you noticed a teen.

BRUMFIEL: For perspective, I imply, galaxies have been thought to kind a couple of billion years after the Large Bang, you realize, given the universe is round 13.7 billion years outdated in whole. However now, JWST is basically testing that speculation.

BARBER: Yeah. And in addition, to place these discoveries in perspective, one other astrophysicist, Priya Natarajan, identified that we’ve got already detected the oldest galaxy, the oldest black gap, simply since JWST began its science operations final yr. So she’s fairly positive we’ll uncover extra record-breakers in 2024.

SHAPIRO: Temperature data on local weather, age data on black holes – it is a yr of data. That is Regina Barber and Geoff Brumfiel from NPR’s science podcast, Quick Wave, the place you may find out about new discoveries, on a regular basis mysteries and the science behind the headlines. Thanks each.

BARBER: Thanks, Ari.

BRUMFIEL: Thanks.

(SOUNDBITE OF MUSIC)

Copyright © 2023 NPR. All rights reserved. Go to our web site phrases of use and permissions pages at www.npr.org for additional info.

NPR transcripts are created on a rush deadline by an NPR contractor. This textual content might not be in its remaining kind and could also be up to date or revised sooner or later. Accuracy and availability might fluctuate. The authoritative report of NPR’s programming is the audio report.