OpenAI Makes Headwaves With Its Own Voice-Cloning Technology

ChatGPT makers OpenAI seem to be on a roll as the company just unveiled its own voice-cloning initiative.

While the project is said to be both useful and intimidating at the same time, it’s only up for grabs to a limited number of the company’s partners, as of right now.

The latest endeavor is dubbed Voice Engine and it’s designed to clone a user’s voice by simply hearing a voice note comprising just 15 seconds. Additionally, the voice that’s replicated is designed to convey all sorts of emotions and intuitive human speech along the way as well.

This makes the whole AI-themed dialogue very realistic in nature as well. Moreover, the firm adds how it was initially curated in the latter part of 2022 to better support the AI tool, ChatGPT, and its inert ability to convert text to speech. But instead of rolling out Voice Engine to the masses, it has opted to take a halt in that regard as perhaps the world is not quite ready, just yet.

The company rolled out a statement on the matter recently and spoke about how they hope to take on a more cautious and careful approach that’s designed to better cater to respective target audiences.

This carries the potential for voice misuse as mentioned in the blog post that revealed several types of Voice Engine in effect.

Translating content : You can use this for translation in tk multiple language. You can clearly see the voice is close to refrence audio in all languages.
HeyGen is one the ealry adopter of this. pic.twitter.com/86V5E5tpQ7

— AshutoshShrivastava (@ai_for_success) March 29, 2024

Speech generated through AI means is quite different from that seen in referenced audio. And that is the scary part because it does come with its fair share of threats online.

We’re talking about examples where it could be placed in the wrong hands and therefore sure technology might be utilized to remove deepfakes and hence can misinform the audience. However, keeping the risks of misuse on the side, the company feels Voice Engine might be a really useful and successful endeavor for society today.

Whatever the case may be, it’s now getting rolled out in users’ direction from all over the globe including a small figure of trusted individuals. Meanwhile, the company is currently boasting about providing reading assistance to a group of school kids. This would serve as translators too as one person’s voice is made use of to communicate in various kinds of languages.

Meanwhile, another case entails giving rise to voice cloning tech for those individuals who lose their voice, which happens to be a concept that’s quite similar to tech giant and iPhone maker Apple.

Furthermore, the company says it is rolling out Voice Engine through preview modes so that partners can benefit, even if they promise never to make use of this technology for reasons dubbed unauthorized like when impersonating the other.

Partners are said to disclose the respective audience and the voices that are being heard by the software as ‘AI-produced’ to ensure complete transparency when using the technology. In the same manner, OpenAI has added watermarking systems to better detect AI-based audio through Voice Engine.

Still, the company is not providing any kind of guarantee about this being rolled out as a bigger release for all. The voice cloning program is a trial and if things go as planned, we shall be seeing it released across the board. But until then, it all depends on how the whole society generates a response to the growing popularity of AI.

Did we mention how technology is already making it hard to differentiate fact from fiction?

Let’s wait and watch for this experiment to roll out and then see what sort of response comes from the company’s partners.

Image: DIW-Aigen

Read next: Google Has Launched a New AI System in DeepMind Called SAFE that is Claimed to be Better than Humans in Fact-Checking
by Dr. Hura Anwar via Digital Information World

Google Has Launched a New AI System in DeepMind Called SAFE that is Claimed to be Better than Humans in Fact-Checking

A new research has found out that Google’s DeepMind can perform better in fact checking than humans. The research was published in a paper and was titled “Long-form factuality in large language models”. In the paper, a fact-checking method called Search-Augmented Factuality Evaluator (SAFE) was talked about and it was found that SAFE breaks down each fact in a document individually and then fact checks it using search on Google.

One of the authors of the study says that SAFE uses Large Language Models (LLMs) to properly break down each fact and first check each fact itself by using its multi-step reasoning process and then also matches the facts with the results in Google Search. To find out if SAFE can be used instead of humans in terms of fact-checking, the researchers tested 16,000 facts with both DeepMind and Humans. It was found that SAFE matched humans 72% of the time. There was also 100 disagreements on facts between humans and SAFE and SAFE proved to be correct 76% of the time.

The paper claims that this LLM can achieve superhuman types of powers but many researchers argue about what ‘superhuman’ really means. Garcy Marcus who is an AI researcher says that he was confused after reading the term ‘superhuman’. He said that researchers are over-hyping the model and this wording doesn’t fit the model. He also added that the researchers meant that SAFE is better than under-paid human fact-checkers. To be really called a superhuman, SAFE needs to compete against many professional human data fact checkers because it is important for having the correct results.

Researchers are claiming that SAFE is 20 times cheaper than human fact checkers. SAFE was also used on many other models like ChatGPT, Gemini, Claude etc to check if these models make factual errors. The results showed that the larger models have less factual errors. But some of the best models also generated false claims which shows that we shouldn’t over-rely on these models for factual information. SAFE was good at identifying those false claims.

SAFE Code has been open sourced on GitHub. The other researchers can use that code to fact-check their work and find any mistakes they made with some facts. SAFE has a long way to go and there is a lot of more work it still needs to compete with humans.

Image: DIW-AIgen

Read next: 34% of Americans Say They've Never Heard of ChatGPT
by Arooj Ahmed via Digital Information World

34% of Americans Say They've Never Heard of ChatGPT

ChatGPT is one of the most hyped up pieces of tech in the world, but in spite of the fact that this is the case, a fairly significant chunk of Americans haven’t heard of it yet. According to a new poll that was conducted by Pew Research, just 23% of Americans who responded to this survey had actually used ChatGPT at this current point in time. That’s a 5 point uptick from the 18% that said the same in July of 2023, but it’s still not as high as you’d expect.

With all of that having been said and now out of the way, it is important to note that adults under the age of 30 were far more likely to have used ChatGPT than might have been the case otherwise. 43% of them said that they have used ChatGPT at some point or another in the past, which is up 10 pounds from the 33% that agreed with this statement in the summer of 2023.

As for adults between the ages of 30 and 49, the proportion rose from 21% to 27% with all things having been considered and taken into account. 17% of adults between the ages of 50 and 64 said that they use it, up from 13%, and even those who are 65 and older reported some level of usage, or 4% to be precise.

It bears mentioning that 34% of the people that responded to this survey stated that they hadn’t even heard of ChatGPT to begin with. This proportion was as high as 36% for Republicans. 38% of all survey respondents were of the opinion that any and all information provided by ChatGPT can’t be trusted, with 10% saying that they trusted some of the information it gave, and just 2% saying that they trusted it a whole lot.

This survey was meant to establish how much election information ChatGPT could provide that would be trusted in any way, shape or form. The low proportion of people willing to rely on it for this all important task just goes to show that ChatGPT isn’t really a trustworthy source of information just yet.

Read next: A Survey Shows that US Adults Are Perceiving Genders Negatively Because of Social Media Content
by Zia Muhammad via Digital Information World

A Survey Shows that US Adults Are Perceiving Genders Negatively Because of Social Media Content

A survey conducted by Morning Consult among 2200 US adults who use social media found out that many users have negative views about gender on social media. Almost all posts on social media are influenced by gender. Even a dinner or color can be defined as girly or boyish. There are also talks about masculinity and femininity on social media. The survey shows that GenZ has many negative views about opposite genders on social media.

30% of the GenZ adults who use social media say that they have started to feel negatively about men and 22% say that they have negative feelings towards women. The things that have influenced their feelings are mostly poor experiences with the gender on social media. In addition to that, many GenZ adults admitted that they see a lot of negative stuff on social media and often doom scroll too.

Even with all this, there is a difference between US GenZ adults and US adults about how much they hate a certain gender even though the feelings of both of them are negative. 10% of the US adults have said that social media has played a great role in shaping their perception about men and women while 15% of GenZ has said that.

The survey also talks about how for GenZ, social media is like their gendered space. Apps like TikTok and Snapchat are more popular among GenZ women while apps like X and Reddit are full of GenZ men. And in these apps, there are a lot of talks about gender and gender biases. A trend on TikTok called the Dating Diaries is getting popular among women where they talk about their bad dating experiences. On the other hand, men on Reddit are demanding rights for men. Misogyny is often a central content theme on Reddit.

62% of the US adults say that negative content on social media is the main reason why most people are perceiving genders this way, not the social media companies. Even if this can worsen the mental health conditions of GenZ, it is nothing that is a threat for which an action should be taken.

Read next: YouTube Shorts Boosts Earnings for 25%+ of Partner Program Channels
by Arooj Ahmed via Digital Information World

Survey Shows Americans Consider Their Cars as Their Family Members and Most of them Want to Buy Electric Cars

A survey conducted by OnePoll on the behalf of Meineke shows that about half of the Americans think of their cars as part of their family. The survey was done among 2,000 Americans and it was revealed that 51% of Americans said that their car is part of their family while 53% said they would keep their car forever with them if they get the chance. For 22% of the Americans in the survey, they have an emotional attachment with their car while 24% said that they have very fond memories in their car.

Six out of ten respondents also had a practical approach and said that their car is reliable(58%) and 45% said that they want to keep their car forever because of the gas mileage their car has given them. 80% of the Americans said that they would like to keep their car as long as it is saving them money. Some of the respondents(46%) said that they will probably use their current car as they cannot afford a new one. Even with the attachment with the current car, 34% of the surveyed car owners said that they will most likely purchase the same model of their car if given the chance. Only 47% of them said that they will probably look for new and latest models, with the majority saying that they will buy an electric car.

45% of Americans said that electric cars are the future and younger people completely agreed with them. 41% are relieved that they have an option of buying an electric car in the future. 59% of the GenZ and 56% of the millennials agreed to it as compared to GenX(37%) and baby boomers(27%). VP of Marketing Meineke Car Care Centers, Christopher Streahle, said that people are adopting electric cars quickly, with the younger generation more hyped up for it than others. As electric cars are good for the environment too, many car owners are considering buying them.

The survey then asked the respondents the reasons why they would buy an electric car. 35% said that they will buy it as they will then no longer have to pay for gas, 33% said that electric cars are good for the environment and 29% said that they create less pollution. 25% would buy electric cars because of tax credits and 17% said that they would buy them because they have quick and smooth acceleration. But Americans still have doubts about electric cars as 66% say that electric cars have very high costs and 41% think that electric cars will increase their visits to the mechanic.

Read net: These Are The Ideal Cities for Digital Nomads, Madrid Spain Takes the Lead
by Arooj Ahmed via Digital Information World

These VPN Apps Are Turning Android Phones into Cybercrime Proxies

A report put out by the Satori threat intelligence team at HUMAN has just published a list of 28 apps on the Google Play Store that can effectively turn Android phones into proxy servers. 17 of these apps are actually VPNs, which people often use in order to circumvent internet blocks or mask their identity with all things having been considered and taken into account. In spite of the fact that this is the case, their phones may actually be getting used to cover up instances of cyber crime.

With all of that having been said and now out of the way, it is important to note that these 17 apps were available free of charge, which might make them more enticing to download than might have been the case otherwise. The apps, VPN or otherwise, are as follows:

• Lite VPN
• Anims Keyboard
• Blaze Stride
• Byte Blade VPN
• Android 12 Launcher (by CaptainDroid)
• Android 13 Launcher (by CaptainDroid)
• Android 14 Launcher (by CaptainDroid)
• CaptainDroid Feeds
• Free Old Classic Movies (by CaptainDroid)
• Phone Comparison (by CaptainDroid)
• Fast Fly VPN
• Fast Fox VPN
• Fast Line VPN
• Funny Char Ging Animation
• Limo Edges
• Oko VPN
• Phone App Launcher
• Quick Flow VPN
• Sample VPN
• Secure Thunder
• Shine Secure
• Speed Surf
• Swift Shield VPN
• Turbo Track VPN
• Turbo Tunnel VPN
• Yellow Flash VPN
• VPN Ultra
• Run VPN

These apps contain a monetizing app by the name of LumiApps. The way this works is that a webpage is loaded in the background using the device’s unique IP address which then retrieves data. The data is then sent to companies. It bears mentioning that the use of this app is in line with GDPR, but it also ended up being used to convert devices into proxies.

Image: HumanSecurity

After this report was published, Google ended up removing any and all apps using this monetization platform because of the fact that this is the sort of thing that could potentially end up stopping this campaign in its tracks. These apps are now available again, although they likely don’t contain LumiApps anymore. Google Play Protect is now capable of detecting the use of this SDK, and the presence of these apps indicates that the SDK isn’t present anymore.

In spite of the fact that this is the case, some of these apps were uploaded from diverging accounts. That seems to suggest that the app publisher is trying to circumvent previous bans, and it remains to be seen what Google will end up doing on the matter at hand.

Any user that is using one of the aforementioned apps would do well to delete them as soon as possible. Failing to do so could result in their devices being turned into proxies at this current point in time.

Read next: Cyberbullying Increased After the Pandemic, New Report Reveals
by Zia Muhammad via Digital Information World

Cyberbullying Increased After the Pandemic, New Report Reveals

A new report by the World Health Organization has just revealed that cyberbullying among school age children increased during the pandemic as well as in its aftermath. Around 17% of the 37,000 youths surveyed in Wales said that they experienced bullying, and overall the survey involved just under 280,000 children, or 279,000 to be precise. These survey respondents hailed from 44 different countries around the world.

With all of that having been said and now out of the way, it is important to note that 12% of boys experienced bullying in 2018, and this number has now risen to 15%. As for girls, the proportion rose from 13% to 16%. The number was even higher in England, with 19% reporting experiencing bullying in some way, shape or form online. This proportion represented the number that had been bullied within the past two months.

It bears mentioning that 11% also acknowledged that they had cyber bullied someone else in that time period with all things having been considered and taken into account. Over in Scotland, the same percentage reported bullying others online, whereas the ones that experienced bullying was at the 18% mark.

Cyber bullying is a serious concern because of the fact that this is the sort of thing that could potentially end up seriously harming the mental health and wellbeing of children. As a result of the fact that this is the case, studies like these are an essential component of the process by which the matter at hand can be fully understood in its entirety.

More work must be done to educate young people about why cyberbullying is something that they shouldn't be taking part in. The pandemic likely drove this trend due to the reason that children were cooped up at home during lockdowns and didn’t really have anywhere that they could go. With screen time frequently hitting six hours per day and the relative anonymity that comes with online interaction, an increase in cyber bullying appears to be something that’s rather inevitable. At the end of the day, it’s something that needs change all in all.

Image: DIW-Aigen

Read next: Cloudflare Blocked 9.3 Million Emails Every Day in 2023
by Zia Muhammad via Digital Information World