I’ve tried many software packages for these modes (there are many, many alternatives!) but for my station I keep coming back to the following: If you aren’t familiar with digital modes, have a look at the brief video below for starters…. The authoritative record of NPR’s programming is the audio record.I spend a lot of time working digital modes, mainly FT4, FT8, PSK31, PSK63, Feld Hell, RTTY and Olivia. This text may not be in its final form and may be updated or revised in the future. NPR transcripts are created on a rush deadline by an NPR contractor. Visit our website terms of use and permissions pages at for further information. Geoff Brumfiel, thank you.Ĭopyright © 2023 NPR. SHAPIRO: From your mouth to the IRS' ears. That researcher I spoke to, Eno Reyes, though, adds that you do not want GPT to do your taxes. And they say, for those reasons, people who use it should be careful. What does OpenAI say about this?īRUMFIEL: Well, they acknowledged that GPT does get things wrong, and it does hallucinate. Not only her byline, but, like, the whole story was made up. SHAPIRO: It put her byline on something that the chatbot wrote?īRUMFIEL: Yeah. Somebody had used ChatGPT to research about, you know, woodworkers and come up with this story that Nurith had supposedly written, but it wasn't real. In fact, one of our journalist colleagues, Nurith Aizenman - she actually got contacted this week about a story she supposedly wrote on Korean American woodworkers, except she never wrote the story. I spoke to a researcher named Eno Reyes, who works for an AI company called Hugging Face, and he told me these AI programs are basically just giant autocomplete machines.ĮNO REYES: They're trying to just say, what is the next word, based on all of the words I've seen before? They don't really have a true sense of factuality.īRUMFIEL: That means that they can be wrong, and they could be wrong in really subtle ways that are hard to spot. I mean, this gets to the real fundamental issue about these chatbots, which is they are not designed to fact-check. SHAPIRO: Why were there errors if this stuff was already on the internet?īRUMFIEL: Right. SHAPIRO: OK, so you're not, like, the next supervillain in the Marvel Universe. And also, they said there were some errors in there. This stuff is already on the internet, which makes sense because that's how OpenAI trains ChatGPT. I gave this to some real nuclear experts, and they said, look. But I should say there's no need to panic. Teller about his work, and I got about 30 pages of really detailed information. But I worked around that by simply asking it to impersonate a famous physicist who designed nuclear weapons, Edward Teller. And so, you know, OpenAI has tried to put in guardrails to prevent people from using it for things like, say, designing a nuclear weapon. SHAPIRO: But you found some problems, like apparently you got it to tell you some things about nuclear weapons that it's not supposed to share.īRUMFIEL: Yeah, I am a big nuke nerd, as people may know. So it definitely seems to be a lot more capable than the previous version. It also, according to OpenAI, passed a bunch of academic tests - several AP course exams - and it has the ability to look at images and describe them in detail, which is a pretty cool feature. The previous version would get things like simple math problems wrong, and this one does much, much better. How good is it?īRUMFIEL: It's really impressive. You've had a chance to try out this version of GPT. SHAPIRO: NPR's science correspondent Geoff Brumfiel has been testing the waters. But can we really trust AI with our taxes? SHAPIRO: That's Greg Brockman, one of the founders of OpenAI, showing off GPT's mad tax skills. It's way, way better than I am at mental math. GREG BROCKMAN: Honestly, I - every time it does it, it's just - it's amazing. Google announced plans to roll out new AI tools across email and its other productivity software, and OpenAI unveiled a new version of its chatbot, ChatGPT, that it claims can figure out someone's taxes. It's been a busy week in the world of artificial intelligence.
0 Comments
Leave a Reply. |