Recently Developed Technology Identifies Prejudice in Advanced Generative Artificial Intelligence Model

As technology advances, new helpful and powerful tools come to the forefront.

A recent example of one is T2I (Text-to-image) generative AI tools that have been adopted by many as they possess the impressive capability to generate images after giving them just a few words as prompts. Text-to-image generates videos and photos that appear so genuine that people may find it difficult to figure out whether they are real or AI-generated. Those generated images and videos offer use in many areas, such as advertising, electoral campaigning, and art.

But prejudices have been reported after viewing the images these tools generated as developers developed them on data taken from humans. These prejudices surround the color of skin and gender.

The tools may be helpful and massive but may also be prone to causing harm to marginalized groups by increasing discrimination and boosting stereotypes.

Researchers from Baskin Engineering at UC Santa Cruz teamed up with Xin Eric Wang, an assistant professor of CS and Engineering, to develop a tool they named, Text to Image Association test. The tool quantifies and identifies those complex prejudices related to humans that the text-to-image generative AI tools have integrated, and the researchers tested it with Stable Diffusion, the advanced generative model. The tool analyzes these biases that encompass various aspects such as faith, ethnicity, gender, and occupation.

Jailu Wang, the leading writer of the paper and CS and engineering Ph.D. student from UCSC, states that both owners and users of the model are concerned about those biases, such as the high possibility of the model only generating images of people from privileged communities to a user from a marginalized community.

The tool that the researchers developed requires the user to instruct the model to generate an image based on a neutral prompt. It could be a prompt related to anything, like a child learning math. After that, the user provides instructions on gender, such as boy learning math and girl learning math’ As a result, the tool compares the disparity between the generated images from the neutral prompt and the specific prompts. Then the tool measures how different these pictures are from the model, quantifying the extent of bias.

Researchers concluded that Stable Diffusion repeats and intensifies these human biases by incorporating them into their generated images. They further tested the tool to draw the connection between two ideas, such as arts and science, along with two characteristics, like female and male. Once it examines it, the tool reveals a score that represents the connection linked to the idea and the characteristic, along with a value to showcase how sure the tool is about the provided score.

To further test the tool’s ability, the team made the model to examine six sets of contrasting ideas by relating them with positive or negative features. The ideas included musical instruments and weapons, insects and flowers, dark skin and light skin, African American and European American, Christianity and Judaism, and gay and straight.

Results showed that the model complied with the harmful stereotypes associated with the concepts but surprised the team by relating dark skin as pleasant and light skin as unpleasant. The model revealed to associate art more with females, science more with males, family more with females, and careers more with males.

This new tool is an advancement after comparing it to a previous technique for assessing prejudice in T21 models, which involved annotating the results after instructing a neutral prompt. The researcher would have to give a neutral prompt, such as a student studying math’ and then identify it according to the image generated by the model, whether it is of a boy studying math or a girl. This technique proved to be inefficient as it was prone to biases relating to gender and was costly.

Xin Wang reported that they aim to replace the annotating technique done by humans with a quick and efficient digital tool that assesses these prejudices. Moreover, the team’s tool also takes note of the background of the generated images, such as the tone and colors used.

Furthermore, the team incorporated Implicit Association Test in their tool, a test to assess stereotypes and human biases. The team hopes that their tool can help software engineers during the development process of their models to address and prevent such biases.

Read next: Google’s AI Ambassador Agrees That AI is a Threat, Claims Google is Different
by Ahmed Naeem via Digital Information World

The FTC Has Received Over 500,000 Credit Bureau Fraud Claims in H1 2023 Alone

Credit bureaus are institutions that can assess your ability to pay back debt and then sell this data to a lending agency with all things having been considered and taken into account. In spite of the fact that this is the case, they can also be involved in a number of scams especially in situations where users are not aware of what they can provide. It turns out that credit bureau frauds are reaching record breaking numbers according to new data.

With all of that having been said and now out of the way, it is important to note that there have been over 517,000 credit bureau fraud claims in the first half of 2023 alone. This comes from information presented by the FTC itself, and it lends credence to the notion that these scams are becoming an ever more prominent component of the threat landscape.

This is almost double the number that was seen last year when these reports seemed like they had started to plateau somewhat. The dip seen in 2021 is also now long gone, and these threats are now on the rise once again as identity fraud becomes rife in the industry.

In 2019, a 18% increase was seen over year. By 2020, this annual growth rate surged to a whopping 89%. 2021 was a much better year, although it too saw a 39% increase overall, with 2022 seeing a 49% increase. If the current trend persists, 2023 may prove to be the single worst year of the decade, and consumers may soon be overwhelmed by these types of frauds.

Untold quantities of data including highly sensitive and personal details such as credit card information could end up in the wrong hands. People committing these frauds often use tactics such as dumpster diving to get the information that can allow them to access accounts. Preventing access to lines of credit can be disastrous not just for consumers but also for the industries that are attempting to cater to them, so more work must be done to keep these threats at bay.

Read next: New Research Reveals that VR is Susceptible To Getting Hacked
by Zia Muhammad via Digital Information World

Dodgy "Beta" Apps: The Cyber Crooks' New Playbook to Snatch Your Crypto

Imagine this: You're strolling through the digital aisles of your favorite app store, searching for the next big thing in cryptocurrency investment. You come across an app that claims to transform your spare change into a cryptocurrency jackpot. You press the download button, eager to ride the crypto wave. You've unknowingly fallen into the latest cyber crook's trap: the "Beta" software deception.

These cunning cybercriminals have discovered a way to circumvent the app store's security by disguising their dangerous products as harmless "beta" versions. You see, beta applications are preliminary drafts of apps created for computer geeks to test and report any flaws. They're like movie trailers: not precisely the whole product, but enough to pique your interest.

The problem? These "beta" apps don't get the usual rigorous security checks. It's like slipping into a movie theater without a ticket; you're in, but no one checks your luggage. As a result, disguised as legitimate, these evil programs pass past the protections and directly into your smartphone.

Consider this: You are lured into a world of bogus cryptocurrency fantasies. The program requests your login information, promises fantastic rewards, and demands money. You eagerly hand over your virtual keys to the kingdom, unaware you're inviting cybercriminals in.

The FBI is on the case, flashing its digital badges as a warning. They're not only ringing the warning but also providing survival advice. "Check the reviews, folks!" they say. Legitimate applications have buddies - genuine people who have tried and tested them. Those with no reviews? They're the digital equivalent of a haunted house – enter at your own risk.

Let's go back to last year when Sophos' tech watchdogs blew the whistle on this operation. They revealed how these cyber swindlers carried out their schemes using Apple's TestFlight system, an innocent tool gone wild. It's like that guy who portrayed a hero in the first film but then became the villain in the sequel. Isn't that sneaky?

Today, these criminals have a fresh hit: the 'CryptoRom' tale. It's like a blockbuster with endless sequels. This time, the bad guys insert harmful code into the program after it's been authorized. It's as though the screenplay was rewritten halfway through production - and not for the better.

Don't think you're secure, Android users. Google's Play Store also has a beta entrance. But are they as relaxed as Apple's bouncers? That's anyone's guess.

So, how can you defend yourself from these online thugs? Easy-peasy. "Trust your instincts," urges the FBI. Something is odd if software requests access to your pet images while claiming to help you invest in cryptocurrency. And keep in mind that malware leaves traces - battery depletion, strange adverts, your smartphone turning into a small oven - you get the idea.

The bad guys are constantly striving for their moment in the spotlight in a world where app stores are the silver screens, and applications are the starring characters. But don't worry, dear reader, armed with reviews, skepticism, and a dash of tech-savviness, and you can enjoy the show without falling for the oldest trick in the cyber book. Happy app searching, and remember to remain cautious to keep safe!

Read next: Reasons Why You Should Not Put Faith in AI
by Rubah Usman via Digital Information World

Telegram's 10th Birthday Bonanza: Stories, Cameras, and Privacy Galore!

Do you remember when you first made your Facebook profile? Probably not, but you certainly remember using Telegram for the first time, right? Hold on to your virtual hats because Telegram is turning a decade old! That's right, a decade of talks, gifs, and sticker-fueled laughing. However, guess what? They're not just blowing out candles but also opening a slew of swanky new gifts for you!

First up are the cool kids on the street, "Stories" Telegram, like the other popular applications you use, now allows you to share your life's tiny adventures with your friends. It's the equivalent of inviting people to your digital show-and-tell arena. Simply drag your contacts down, and you're done! Stories are right at the top, chillin'. You can react, reply secretly, or sprinkle some emoji magic – it's your virtual stage, superstar!

Now, let's speak about a camera trick that will give your selfies and photos a hip new look. Folks, it's Dual Camera Mode! Consider this: you're in both selfie and regular mode at the exact moment. One camera is like, "Hey, check me out," while the other is all, "I'm the star!" You may choose which camera receives the spotlight and even shift it around. On your screen, it's like a camera dance-off!

But wait, there's more: you'll get VIP service! Welcome to Granular Privacy Settings. It's like being in charge of your story-sharing world. You decide who gets a backstage pass to your Stories – everyone, just contacts, close pals, or your hand-picked dream team. No more oversharing mishaps – you're the gatekeeper!

And if you're nostalgic, Stories in Profiles are your new BFFs. You may keep those great stories on your profile for as long as you wish - forever, people! It's similar to catching fireflies in a jar but much more computerized and less... firefly-ish.

Editing Stories is a tech-savvy magic trick. Yes, you can make modifications on the go without starting from scratch. Change captions, stickers, and even visibility like a narrative magician. There will be no more erasing or drama - just pure editing enjoyment!

"Who's watching my stories?" you may be asking. Now you can stop wondering. Detailed View Statistics is here to help. You get the inside scoop on who is looking over your work.

Sort them by reactions or recent peeps – it's like a backstage pass to the showbiz of your stories.

We have a hidden weapon for VIP friends out there: Stealth Mode. Have you ever seen something that you wish you hadn't? Poof! You can erase your tracks faster than a ninja on roller skates with Stealth Mode. Sneak in and out - your opinions, your rules!

The celebration has already begun on Android and iOS. If you're experiencing FOMO, simply visit your app store - it's like a treasure hunt for new features. Don't worry if you don't see the update button yet. Like a digital fairy godmother, your app will update when the moment is right. So, there you have it, fellow Telegram adventurers – swipe, share, and celebrate the telegram-tastic journey!

Read next: Android And iOS Witness 24% Rise In In-App Purchases During H1 2023
by Rubah Usman via Digital Information World

48% of Teenage Girls Receive Unwelcome Advances Through Messaging Apps

Instant messaging has completely and utterly transformed the face of the world because of the fact that this is the sort of thing that could potentially end up facilitating speedy communication. However, it is also seen as a bit of a double edged sword with all things having been considered and taken into account. For example, a recent survey conducted by Common Sense Media revealed that as many as 48% of teenage girls in the US got unsolicited messages through these apps.

46% said the same about TikTok, whereas YouTube is marginally better although it too had 30% of teenage girl users receiving unwelcome advances through messages. With all of that having been said and now out of the way, it is important to note that 45% of the teenage girls said that they are addicted to TikTok, which might make this problem even worse than might have been the case otherwise.

It bears mentioning that many of the girls included are between the ages of 11 and 15. This seems to suggest that predators and child groomers are attempting to take advantage of instant messaging apps as well as various social media platforms to find victims. Many of these teenage girls are unable to remove themselves from the apps that are putting them in harm’s way due to the aforementioned issue of addiction which doesn’t seem to be getting better anytime soon.

What’s more, Instagram turned out to be the worst app of all. 58% of adolescent girls on the platform said that they received unexpected messages that made them feel uncomfortable, with Snapchat following at a close second with 57%. In spite of the fact that this is the case, Instagram was cited as addictive by just 33% of girls, with Snapchat receiving 37%.

All in all, the problems of social media addiction and unwelcome solicitations seem to go hand in hand. It will be interesting to see how this impacts policy surrounding social media and screen time in the future, although the responsibility ultimately lies with the parents to do the right thing to the best of their abilities.

Read next: New Research Reveals that VR is Susceptible To Getting Hacked
by Zia Muhammad via Digital Information World

Reasons Why You Should Not Put Faith in AI

When Alexa, Amazon’s AI voice assistant, is asked about Amazon being a monopoly, it claims not to know. While it quickly criticizes other tech monopolies, it remains quiet about the wrongdoings of the company that developed it. Alexa lies its loyalty to its creators rather than its users. However, in most cases, it’s not as clear whose interests AI systems serve. We must adopt a cautious mindset when approaching AI to protect ourselves from potential exploitation by these models.

This involves carefully crafting the input we provide and skeptical evaluating its response.

It has become difficult to distinguish who advantages since newer AI systems give more complex and reduced memorized responses. By now, we are aware of websites and platforms using dark patterns to make you keep using their services for longer durations, which profits them through ad revenue while dismissing your welfare. Google, TikTok, and Facebook are all a part of it.

But what sets AI models and these online websites apart is the way they interact, which has started to take the shape of real-life relationships, as it is easy to predict that in the near future, AI will be planning travels and negotiations for you, and possibly serve as mental health counselors.

Generative AIs are already on the track to become 24-hour digital assistants that you can customize to your preference. ChatGPT is an example.

Data scientists and security experts have claimed that people who depend on AI will unknowingly trust them to help plan their days. Ensuring that AIs are not clandestinely serving other individuals is essential. People are unaware that the platforms and devices they use may not be operating in their favor. Apps may sell personal data without your permission, and devices with cameras may spy on you. There is no way to ensure security when surveillance capitalism uses AI for its benefit and is quiet about its doings.

To use AI efficiently, you need to let it know everything about you, including things that people close to you know about. And sometimes, things that only you are aware of about yourself.

Digital giants have funded AI developers in large amounts yet offer their services to users for free or for a nominal price. How do these tech monopolies monetize from all of it? The answer is the same way that most online services make money, which means manipulation and surveillance are involved.

If you ask AI to decide where you should eat on your holiday, it will pick a restaurant that probably has incentives for its developer. You must have noticed that with paid Google search results and paid ads on Instagram and Facebook.

You may not also know if the AI model that you’re using generates an unbiased response when asked about government updates. There is no way of telling if the generated responses support the owner’s political alliance. Nor can you tell if the candidate has bribed them for support.

We should expect these AI companies to do better for people to trust their services. The proposed AI Act by the EU took a significant step in demanding transparency in data used for AI training, addressing potential bias, being open about possible risks, and providing reports on industry-standard tests.

The government should intervene to take initiatives to protect consumers of these AI models. Until then, users would have to use AI at risk and approach their responses critically.

Read next: Google’s AI Ambassador Agrees That AI is a Threat, Claims Google is Different
by Ahmed Naeem via Digital Information World

Android And iOS Witness 24% Rise In In-App Purchases During H1 2023

The latest stats from leading software company Tenjin are out regarding H1 2023 and there were some clear shakeups worth a mention.

Most people usually consider drops in eCPMs as a measure of declining success as they take it as the way through which monetization of digital content takes place. Recent stats proved that while the eCPM fell by 26% across Android, we saw a smaller decrease of 12% occur for iOS during H2 of last year and H1 of this year.

Remember, ad monetization continues to stay stronger than ever and is a crucial component of today’s digital landscape. See, you’ve got to create a balance between the different types of models that are up for grabs in terms of monetization. Only then can you expect to control the whole landscape better.

Meanwhile, the report was seen highlighting the fact that there was a major rise in the figure for purchases done through apps of both Android and iOS, where a growth of 24% was seen.

That’s truly great news and proof of how game developers are really making the most of all the hybrid monetization opportunities out there today, implementing each as it comes their way.

Coming down to regional differences, the report failed to highlight too much of a change as far as the leading nations are concerned in terms of combined app installations for Android. But for iOS, we’re seeing the UK overtake the likes of China, Germany, and even Japan.

There are so many changes arising in the landscape for iOS and that’s resulting in more app developers and even ad providers to better their business models.

Seeing the United Kingdom take center stage as a leading user hub and possible source of revenue means saying hello to advertisers that wish to customize campaigns that are in line with UK users’ choices. And by altering such monetization strategies, it may even attract the upcoming user base in the country which is a signal for greater subscriptions and more purchases done on the app.

Such changes are also putting into perspective the significance of comprehending what the locals do prefer in terms of ads and monetization for apps. Remember, as the United Kingdom’s uprising occurs, so many other countries may follow in its footsteps with greater app engagement.

Remember, more developers and those putting out ads could remain flexible and would also be open to change as they take user behavior into consideration from across the board.

Read next: New Study Claims YouTube Is More Popular Than TikTok Among Those In Search Of New Book Reads
by Dr. Hura Anwar via Digital Information World