The Humane Newsletter #30
The 30th edition of our Humane Newsletter is out, wow!
And it's as interesting as ever, speaking about AI effecting our beliefs, predicting income levels through social media posts, and a privacy focused browser.
Check it out below!
Or read it and all our previous editions by clicking here - for COSE.
Also, we've got some exciting news coming, so follow us to stay posted!
Misinformation Express: How Generative AI Models Like ChatGPT, DALL-E, and Midjourney May Distort Human Beliefs
Living in a high information world, where we can access answers to more or less any and every question, the fear of miss- and false information has become a hot topic and a growing issue. Instead of using the news platforms we previously relied on, or even search engines like Google, there’s been an increase in using AI chats to find answers to any and every kind of question.?
Now, researchers are warning us that these generative AI models could distort human beliefs by spreading false information, maintain and spread stereotyped biases. Researchers Celeste Kidd and Abeba Birhane argue that the capabilities of AI, through society’s perception and expectation, has been exaggerated.?
There’s a widespread belief that these AI models surpass human abilities, which leads to believing the information handed to us by these models without further investigation or questioning. And we, as individuals, tend to be more susceptible to influence when seeking information.?
Scientists Accurately Predict Individuals’ Income Solely off Social Media Posts
A study focusing on the social networking site Nextdoor shows that scientists were able to accurately predict user’s income, only through using the content of their posts.?
领英推荐
This could mean that users' income can be predicted throughout our overall online activities, or at least give insight into our socioeconomic standing. In turn, this enables the potential for user profiling, paving the way for social media platforms to be even more specific with suggested content.?
But also advertisers and e-commerce businesses could be taking advantage of economic profiling to target consumers more effectively, by promoting products at varying prices based on the user’s income level.?
The study also showed a difference in the types of post between wealthier and poorer neighbourhoods, showing that the wealthier ones were more likely to share positive posts but would also discuss crime - despite the crime rates being lower than in the poorer neighbourhoods.?
DuckDuckGo browser beta for Windows bakes in a lot of privacy tools
Many of us value our privacy, however on the web we’re not as private as many of us might believe. Every internet website with ads identifies you as soon as you press the site, through analysing your data the ad slot for that website will be directed specifically to you labeling you based on the data as ‘your interests’.?
As a result, this way of advertising has been and still is increasing, but with us as users becoming aware of the lack of online privacy and anonymity, there is an increase in initiatives for change.?
DuckDuckGo is a privacy-focused firm who now has released a public beta of its browser for Windows after debuting for Mac, over a year ago.?
The browser has enforced encryption, Duck Player with many YouTube videos without privacy-invading ads and without constantly giving you recommendations, cookie pop-up management, email protection and more privacy focused features. DuckDuckGo claims that its browser uses less data than Chrome by pre-blocking trackers, however the browser still lacks many common extensions, having the basics of password management and ad-blocking tools, as well as Duck Player.?
DuckDuckGo has been relying on Microsoft or Bing for most search results and advertising, but product director Peter Dolanjski points out that the browsers engineers have spent time addressing privacy issues, making sure that crash reports aren’t sent to Microsoft.?