Top AI Companies Commit To The White House’s Initiative Of Curbing Creation And Distribution Of Explicit Images

The White House has recently confirmed that a number of top tech giants are more than eager to curb the issues related to AI misuse. This includes the problem linked to producing and distributing explicit images.

The list includes Adobe, OpenAI, Common Crawl, Anthropic, Microsoft, and Cohere. All those participating in the cause also rolled out the measures they’re taking to stop their platforms from being used for this unlawful act. This includes the generation of nonconsensual pictures of adults and children which are classified as NCII and CSAM.

The issue is nothing new and has been gaining a lot of negative publicity for so long, especially since the trend of generative AI took center stage. The tech giants claimed to be committed more now than ever to safeguarding users from such acts that’s sexual abuse.

All the companies other than Common Crawl also mentioned how they would add feedback loops and stress-testing maneuvers to developmental processes. This is to protect against any AI models rolling out sexual abuse images. It would similarly be deleting nude pictures made through AI datasets when applicable.

The commitment is a serious one and from what we can tell from today’s announcement, it’s not any different from what we’ve heard from the past. All companies have made similar ones but no laws were outlined to action those that don’t stay true to their promises. Still, the fact that they do care and consider the need for change is a great step forward.

Meanwhile, it’s not hard to see that some tech giants were missing from the list including Meta, Apple, Google, and Amazon. These companies have been slammed since day one for not doing enough and not being on the list sends out a loud and clear message.

We’ve seen several big AI and tech firms make efforts for victims in the past that were impacted by non consenual images produced of them in a sexual manner. Their goals were linked to stopping their distribution, other than the usual federal way.

We’ve seen companies like StopNCII collaborate with other leading tech giants for a more impactful approach to deleing such content from databases. Meanwhile, other companies are committed to rolling out new tools that report these kinds of deepfakes and sexual images of users made without consent.

Image: DIW-Aigen

Read next: OpenAI Says Its Latest ‘o1’ Model Excels In Complex Reasoning And Performs Better Than Humans In Math, Science, And Coding
by Dr. Hura Anwar via Digital Information World

OpenAI Says Its Latest ‘o1’ Model Excels In Complex Reasoning And Performs Better Than Humans In Math, Science, And Coding

OpenAI is making some extraordinary claims about its newest language model o1.

The company was seen announcing some breakthrough advancements in human reasoning and also outperforming the average human mind in certain domains. This includes math, coding, and also science.

The biggest advantage spoken about was certainly complex reasoning. This came in the form of a new blog post released by the AI giant who says it’s at the same level of human performance, if not better. This is the first time we’ve seen OpenAI claim such features but the actual benefits are still speculative.

For instance, it was hailed as scoring in the 89th percentile across various programming challenges which are brought out by Codeforces. Similarly, the firm talks about the model performing at the same level that would ensure it was amongst the top 500 students performing at the top of the AIME.

Furthermore, the company mentioned how o1 goes above and beyond the average performer in human subjects including those with PhD on a combination of chemistry, biology, and physics examinations. Wow, that’s certainly a major accomplishment, if that’s really the case.

However, experts are skeptical as testing in real-time is yet to be done. And until and unless that occurs, it’s hard to confirm.

Other than that, reinforced learning was discussed in detail including how o1 can simplify even the most complex equations through a chain of thought mechanism. This is done through a step-by-step maneuver where mistakes are rectified and adjusted before a final reply gets generated.

OpenAI also spoke about how the o1 model developed the best reasoning skills when compared to standard models out there today. But it’s still quite unclear how claimed reasoning might better comprehension of complex questions throughout certain areas like coding, math, science, and beyond.

When you look at things from the SEO perspective, content interpretation is enhanced and the capability to answer queries directly might be useful. It’s so wise to be careful and see testing done through third parties.

Remember, OpenAI needs to go above and beyond the classic benchmark that entails browbeating and rolling out objectives. Remember, the world needs evidence to understand if these claims are true or not. Adding capabilities to the AI model can better display realistic uses. What do you think?

Image: DIW-Aigen

Read next: Gamelight Effect: Setting the Standard for Success in Rewarded UA
by Dr. Hura Anwar via Digital Information World

Gamelight Effect: Setting the Standard for Success in Rewarded UA

Gamelight has emerged as a game-changer in the mobile gaming industry. Through its fresh take on rewarded user acquisition (UA), it has turned a once-overlooked tactic into a central driver for user growth. Gamelight's success has sparked a trend in the industry. Seeing the potential, startups and established companies are quickly introducing their own rewarded products, changing the way user acquisition is done.

Gamelight's Distinctive Strengths

Gamelight has revolutionised the rewarded UA landscape by deeply integrating user behaviour analysis, app usage data, and demographic insights to create comprehensive user profiles. These profiles enable the platform to deliver highly personalised game recommendations that align perfectly with individual preferences, significantly boosting user engagement and satisfaction. Furthermore, Gamelight’s platform rewards players by offering bonuses for extended gameplay and achieving milestones, which can be redeemed for gift cards and vouchers—strategies that not only enhance retention but also build strong player loyalty.

The platform’s AI-driven targeting ensures that users are presented with games they are most likely to enjoy, which results in improved retention rates and higher lifetime value (LTV). By focusing on key performance metrics such as ROAS, LTV, and long-term retention, Gamelight consistently delivers excellent results, giving its partners a distinct edge in both user acquisition and revenue growth.

What truly sets Gamelight apart is its ability to deliver users at scale. In a short time, it has become the leading rewarded channel in the industry, quickly surpassing its competitors.

Driving Industry-Wide Innovation

Gamelight's groundbreaking success has spurred a wave of innovation, inspiring both new startups and established ad companies to develop their own rewarded products. These companies recognised the potential of the rewarded UA model and have contributed to reshaping the industry landscape. As more businesses enter this space and innovate, the entire ecosystem benefits, leading to further growth and new opportunities.

Additionally, Gamelight has played a crucial role in expanding iOS's presence within the rewarded UA space. Previously, iOS had a minimal role in this area, but Gamelight’s influence has led to broader adoption, with more companies adopting its product model and integrating similar approaches into their offerings.

Rewarded Channels: Redefining the UA Landscape

Gamelight’s success has significantly shifted the role of rewarded channels in mobile gaming. Rewarded UA has now become a primary focus for many companies, with substantial portions of their marketing budgets allocated to these strategies. This transition is driven by the high-quality users and superior retention rates that rewarded channels offer compared to traditional methods. As a result, companies are able to attract more dedicated and engaged players, establishing new benchmarks in user acquisition.

Gamelight's innovative approach has transformed rewarded UA from a niche practice into a fundamental aspect of mobile game marketing. By redefining how game publishers interact with users, Gamelight has set new standards for retention and lifetime value (LTV). As the rewarded UA model gains traction, Gamelight continues to lead the way, driving success for its partners and shaping the future of mobile marketing through ongoing innovation.

Conclusion: The Path Forward for Rewarded UA

The landscape of rewarded UA has changed significantly thanks to Gamelight's groundbreaking innovations:

• Rewarded UA has grown into a vital strategy for mobile gaming, largely due to Gamelight’s innovative approach.
• Gamelight’s AI-driven platform has revolutionised user engagement for game publishers, resulting in higher retention rates and lifetime value (LTV).
• The success of Gamelight has sparked an increase in rewarded apps, validating the model and expanding the ecosystem.
• Gamelight views this growth as a positive trend, one that encourages ongoing innovation within the industry.
• As the rewarded UA sector continues to expand, Gamelight remains at the forefront, shaping the future of mobile game marketing and driving success for its partners.
• With its ongoing advancements, Gamelight is well-positioned to continue leading the evolution of the rewarded user acquisition landscape, setting the stage for future industry breakthroughs.

by Web Desk via Digital Information World

Study Shows Privacy Policies on Different Tech Platforms Give a Hard Time to Advertisers to Manage their Ads

According to a new study from the University of Notre Dam, tech companies that remove their user data after some time are a big threat to advertisers. One of the examples of this is Google who started removing their user data if their account was inactive for 18 months. This was a move to preserve user data and privacy but this type of practices can limit ad variety. The study looked at how the designs to protect users’ privacy can have negative effects on the advertising industry.

The study focused on Multiple-Purpose Ads (MPAs) where different ads are present within the same space. When companies protect their user data, it results in fewer user engagements on ads and the click through rate of ads also reduces. Due to this, the ad varieties also decline and users cannot find relevant ads in MPAs. It is because the time ad companies get to target consumer data becomes shorter. When the companies look for user data in a short time, they cannot find many specific categories to target the users with relevant advertisements.

For instance, if a user has searched for men’s shoes, hiking and basketball within 14 days, advertisers won’t know which types of ads to show. If they show the ads in the same categories, the consumers will lose interest. If they get the exposure of different ads, there is a chance that it will strike consumer interest. The author of the study said that these kinds of things are important to consider if advertisers want to balance consumer privacy and effective advertising strategies.

The study looked at two additional policies which are related to ad variety. The first one was the economic impact of ad policies and how advertisers can incorporate clicks and bids according to consumer interest. The next one was how researchers can modify the reservation price in ad auctions. By making use of both factors, advertisers can make their ads effective and increase their revenue on different platforms. This research shows how advertisers should maintain a harmony between user privacy and ad effectiveness.

Image: DIW-Aigen

Read next: Tech Jobs Seen as Best for Family Balance, Survey Finds 17% Prefer Their Flexibility
by Arooj Ahmed via Digital Information World

Meta Under Fire After Australian Government Questions Company About Scraping Users’ Data For AI Training

Meta’s Privacy Director is under fire after a new inquiry by the Australian Government spoke about its data collection practices in the country.

Surprisingly, it was found that no Meta user was given the option to opt out of data-collecting practices. Now, the tech giant is asked why users’ data was scrapped to train its AI models.

As per reports from ABC News, Melinda Claybaugh denied all the claims at the start but after being pressurized to give more answers, she admitted that the company does indeed take part in scrapping all images and texts from social media posts that go as far back as the year 2007.

The only condition where that may not have taken place is if posts were put on private settings. Furthermore, she did admit how the organization is not offering Australians any option for opting out at the moment, similar to the benefits allocated for those in the EU.

She also confirmed that Meta does not take part in scraping user accounts of those below the 18-year age group. However, she did mention another alarming point on how it collects images and other data if it's posted to their parents’ accounts.

Then came the tricky part where she was asked why Australians don’t get any option for consenting to data collected. To that her response was vague, adding how the EU one comes as a reply to legal regulations. This is likely to be linked to the GDPR.

Meta did notify several users across the EU about how it was collecting their data for the sake of training AI unless they checked the opt-out feature.

While Meta might have agreed to the EU’s framework, she still thinks the legal query linked to interpreting any privacy law related to AI is questionable. Meta also opted to avoid rolling out new multimodal AI models and future variants as it feels unclear about what regulators really want.

A lot of issues are related to the issues that have to do with training AI models and using data from EU users while still being in line with the GDPR.

Despite legal queries related to adopting AI in the EU, the conclusion is that Meta is providing EU users more benefits than those in Australia. The reason is simple, they have no other option. Only if such a law is present in Australia can users benefit from similar functionalities.

Now is that really fair? To brush off users’ privacy rights for a poor justification like that does not seem like Meta is winning votes from policymakers in the country, any time soon.

The government has failed big time in terms of acting on privacy and therefore seems to be more focused on monetizing and exploiting the data collected by users. This includes minors on its Facebook app. What do you think?

Image: DIW-Aigen

Read next: Data Shows Consumers Have Lost $5.6 Billion in Cryptocurrency in 2023
by Dr. Hura Anwar via Digital Information World

Data Shows Consumers Have Lost $5.6 Billion in Cryptocurrency in 2023

According to a new report by the FBI, $5.6 billion in cryptocurrency were lost to criminals in 2023, which is a 45% increase from 2022. It is the first time that the FBI has shared a report about cryptocurrency. The FBI received more than 69,000 complaints about cryptocurrency fraud last year, and this makes only 10% of total financial fraud complaints to the FBI.

In a call with a journalist, the FBI said that the reason why complaints about crypto fraud have increased a lot is because many people are no longer hesitant to report crypto crimes. The other reason is that many people believe that they can become rich by investing in crypto, including criminals. So other types or crimes, like ransomware and hacking are also related to cryptocurrency somehow. According to the report, most criminals stole cryptocurrency through investment fraud, with $3.96 billion lost last year. There were approximately 5200 reports in each age range (30-39 and 40-49) which reported investment frauds in cryptocurrency.

Most criminals do investment frauds by fake apps or sites which look real, and lure customers to invest in crypto. Slowly, they look into users’ funds and then steal the money from them. Another crypto fraud quite common nowadays is pig butchering in which criminals strike a romantic relationship with a victim and then ask them to invest in crypto after they gain their trust. Pig Butchering scam is often done by sending the victim a message, email or direct message on social media and then they slowly try to befriend them. It takes many weeks and months to befriend a person or do any romantic advances at them, but once the victim becomes comfortable, they bring out the talk of investing in crypto.

The report also noted that victims can get scammed more than once. If they have lost their crypto money to scammers, victims immediately look for some platforms that can help them recover their funds. So the fraudulent crypto businesses give the surety to victims that they will be able to help them recover their cryptocurrency but take more from them. The FBI asks all the cryptocurrency traders to always look out for scammers and only do an investment if they completely trust the person or have met them in person.

Read next:

• Is Apple Delaying the Obsolescence of iPhones to Prevent Losing Customers to Android Rivals?

• Generative AI Helps Computer Science Students Excel in Software Development

• Tech Jobs Seen as Best for Family Balance, Survey Finds 17% Prefer Their Flexibility
by Arooj Ahmed via Digital Information World

Is Apple Delaying the Obsolescence of iPhones to Prevent Losing Customers to Android Rivals?

Whenever there's a new iPhone launch, Apple drops the support of some of its old iPhones. This week Apple has launched its iOS 18 system and iPhone 16 models. But this year, Apple is not fully letting go of its old iPhone models. One of the iPhone models Apple is continuing to support is iPhone XR. This model was launched in September 2018, which was six years ago. Apple has announced that it is going to support its 6 years old model for another year.

Apple is known for making its old models obsolete by taking away their software support on them every year. Last year, Apple made three iPhone models obsolete and there were more iPhone models made obsolete the year before. This was advantageous to Apple as users would then opt for newer iPhone models by paying about $500 more. There's a term for it known as “upgrade cycle” and it is a favorite of many investors.

This year, however, Apple has decided that it's not going to make any older iPhone models obsolete and will shift their operating system from iOS 17 to iOS 18. This makes about 28 iPhone models that Apple is supporting. Why is Apple doing this if it can earn revenue by making the models obsolete? The answer is that Apple strongly takes its customers seriously. If it makes iPhones obsolete every year, it isn't going to do good to their customers.

Apple is also working on artificial intelligence on its newer models but it won't be available till 2025. On the other hand, Samsung and Google are already selling smartphones with AI features in the market. If Apple stops its support on older iPhone models, their customers can start looking for smartphones from other companies. Losing even a small number of customers to Android would not be good for Apple. Apple has been gaining customers for some years now and it doesn't want to do anything that can make the company lose customers.

There are many Android handsets available in the market and Android companies are promising their customers that they will provide them with the latest device support. Samsung has said that it will provide support to its galaxy models for seven years now and that's Apple's plan too. Apple is just waiting for its Apple Intelligence to come to the market as it will be the biggest upgrade cycle in Apple's history.

Image: DIW-Aigen

Read next: Apple’s Launch of iPhone 16 May Be A ‘Golden Upgrade Cycle’ But The Company Failed To Address Plans For China
by Arooj Ahmed via Digital Information World