The Download: Fact-checking AI and chatbots that debunk conspiracy theories
This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.
Chatbots can persuade people to stop believing in conspiracy theories
The internet has made it easier than ever before to encounter and spread conspiracy theories. While some conspiracy theories are harmless, they can cause discord, and sometimes even lead to deaths. Researchers at Cornell University and MIT Sloan found that chatting with a large-language model (LLM), which simulates human speech, reduced the belief of people in conspiracy theories by 20%. This was true even for participants who said that their beliefs played a significant role in their identity. Read the full story.
–Rhiannon Williams
Google’s new tool lets large language models fact-check their responses
The news:
Google is releasing a tool called DataGemma that it hopes will help to reduce problems caused by AI ‘hallucinating’, or making incorrect claims. It uses two methods to assist large language models in fact-checking their responses and citing their sources to users more transparently. What’s next?
If it works, it could prove to be a boon for Google’s plan to integrate AI deeper into their search engine. It comes with some caveats. Read the full story.–James O’Donnell
Neuroscientists and architects are using this enormous laboratory to make buildings betterHave you ever found yourself lost in a building that felt impossible to navigate? Building design should be centered on the people that will use those buildings. It’s not an easy task. There are many different people with different bodies and minds, as well as varying needs and wants. How can we take into account all of these factors?
To find out, neuroscientists are collaborating with architects in an East London laboratory that allows them to create simulated worlds. Read the full story.
–Jessica Hamzelou
This story is from The Checkup, our weekly biotech and health newsletter.
Sign up
to receive it in your inbox every Thursday.
The must-reads
I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.1 OpenAI has released an AI model with ‘reasoning’ capabilitiesIt claims it’s a step toward its broader goal of human-like artificial intelligence. The Verge)
+ This could be particularly useful for math tutors and coders. Why is it important that AI be good at math? (MIT Technology Review)
2 Microsoft wants to lead the way in climate innovation
While simultaneously selling AI to fossil fuel companies. The Atlantic $)
+ Google and Amazon, the problem with Big Tech’s climate claims. The FDA has approved Apple AirPods for hearing aids. (WP $ )
+ This could change the way people access hearing enhancement devices. The Verge
4 According to Nick Clegg of Meta, its global affairs chief, parents are not using Meta’s child-safety controls. The Guardian)
+ Many technology executives limit their children’s exposure to tech. (The Atlantic $)
5 How AI is turbo boosting legal action
Especially when it comes to mass litigation. False ads offering free money were used to target low-income Americans
Some victims’ health insurance plans were changed without consent. (WSJ $)
7 Inside the stratospheric rise of the ‘medical’ beverage
Promising us everything from glowier skin to increased energy. (Vox)
8 Japan’s police force is heading online
Cybercrime is booming, as criminal activity in the real world drops. Bloomberg $)
9 AI is able to replicate the handwriting of your deceased loved ones
For many, it’s an emotional reminder. (Ars Technica )
+ A new technology allows us to “speak” with our deceased relatives. Are we prepared? (MIT Technology Review)
10 Crypto creators are resorting to dangerous stunts for attention
Don’t try this at home. Quote of the Day
(Wired $).
You can’t talk to James the AI bot.
You won’t be able to have a conversation. The AI hype is built around high test scores. Those tests are flawed.
August 2023
In the past few years, multiple researchers claim to have shown that large language models can pass cognitive tests designed for humans, from working through problems step by step, to guessing what other people are thinking.
These kinds of results are feeding a hype machine predicting that these machines will soon come for white-collar jobs. The problem is that there is little consensus on the meaning of these results. We can still enjoy nice things.
We have a place to relax, entertain and distract you.
An oasis of comfort, fun and distraction that will brighten your day. You can still have nice things.
A place for comfort, fun and distraction to brighten your day. Send me an email or tweet me your ideas.
+ Pearl the Wonder Horse is not just a therapy animal–she’s also an accomplished keyboardist.
+ Pearl the Wonder Horse isn’t just a therapy animal–she’s also an accomplished keyboardist.
+ We love you Peter Dinklage!
+ Money for Nothing sounds even better on a lute.
