Meta Will Not Sign EU’s AI Pact For Sharing Detailed Summaries Of Data Used To Train AI

Tech giant Meta seems to have had enough of the stringent regulations of the EU. The company just took a bold decision by refusing to accept the AI Pact, at least for the time being.

This was first discussed in May when lawmakers spoke about a new pact that governed the use and creation of AI. The goal was to ensure AI was in check with tech giants providing detailed summaries of all the data used to train AI models.

Now, Meta has decided to opt out of the law, giving rise to a new debate in the region.

In the past month, the EC mentioned how most of the rules regarding this law will be applicable starting in August of 2026. But until that happens, a temporary pact was launched that tech giants could sign.

This would outline the important obligations of the rules enlisted.

As per Meta’s representative, the company is open to discussions and working in harmony with the EU but joining the pact right now is not on the cards. With that being said, a possible signing in the future was also possible.

We can confirm that this AI pact is actually the fifth pillar of EU legislation on this front. It is aligned with the Digital Markets Act Digital Services Acts, and even the Data Act in that part of the world.

Image: DIW-Aigen

Read next: The Story of Alexander Wang Who Became One of the Youngest Billionaires in the World from His AI Labeling Startup, Scale AI
by Dr. Hura Anwar via Digital Information World

Google has Completely Disabled Cache Feature, Users Won’t be Able to Access it in Search Results

Google had previously announced that it will remove Google Cache completely from search results. At first, Google removed the cache link and then added the links to the Wayback Machine. Now Google has completely disabled Google Cache by disabling its direct link. If the users go directly to the Google cache by searching for it on the search bar, Google won't show anything in the results, such as this one.

Google Cache has completely stopped working in the past 12 hours. This has made people discuss it on different social media apps and forums. They are discussing how even though Google has completely disabled cache, new pages can still see it temporarily.

Google Liason, Danny Sullivan, said that it's a sad thing that Google cache isn't available but it is something that needed to be done. It was one of the oldest features of Google and people back then relied on Cache when they couldn't open a page because of page loading. Now things have improved a lot so there was no need for Google Cache. Sullivan added that he thinks that Google should add cache links to internet archives within About This Result. This would allow people to see how a page changed overtime.

If someone has a Search Console, they can see what the crawler saw looking at their own page by using URL inspector. The alternatives for Google Cache are Wayback machine, URL Inspection Tool and Google’s rich result testing tool.

Read next:

• Google Introduces Sharper Satellite Imagery For Its Maps and Earth Feature

• Rising Voice Cloning Scams Underscore Urgency for Stronger Public Safeguards and Fraud Prevention

• The Story of Alexander Wang Who Became One of the Youngest Billionaires in the World from His AI Labeling Startup, Scale AI
by Arooj Ahmed via Digital Information World

The Story of Alexander Wang Who Became One of the Youngest Billionaires in the World from His AI Labeling Startup, Scale AI

Alexander Wang, at age 27, has become one of the world’s youngest billionaires. He succeeded in doing so by having more than 100,000 employees on contracts who help him power the modern AI. His startup named Scale AI has hired workers from all over the world and these workers wrote and summarized texts as well as label images for AI to better understand human patterns. This is known as data labeling and the employees summarize news articles or write stories in Hindi to Mandarin.

This data labeling business has become so in demand by Gen AI companies that Scale AI’s revenue jumped to triple last year and now has a value of $14 billion. Wang says that his company’s importance is all due to the computing chip they are using which is developed by Nvidia. Scale AI’s office is located in San Francisco and the company hires individuals through advertising on LinkedIn and Reddit. The work is remote and most of the workers do not know that they are working for Scale AI in their initial days. The employees do their tasks on two websites known as Outlier and Remotasks on which no affiliation with Scale AI is mentioned. Projects are also given code names so the employees do not know what clients they are working for.

Many remote employees get frustrated with their jobs and quit. Sometimes they get delayed payments or hourly wages as low as $8. Some also do not produce quality content and some cheat to increase their productivity. Wang also doesn't like the way his company is working right now. He calls this work mundane and boring and has said that his company is trying to develop new software and AI products.

Scale AI has also worked for Meta and has given the company code word “Flamingo”. Last year, Meta had asked the company to create 27,000 questions and answers so they can train their AI models. When Meta received the data, it turned out that contractors had used ChatGPT to write all the responses. Upon contact with the company, Wang tried to save the project and asked its employees to rewrite everything and send it to Meta. Wang also discovered that many of the badly written data had come from Kenyans so Wang restricted all their projects to the USA and English speaking countries. But it wasn't very useful because many USA accounts were being sold to people from other countries on Facebook and WhatsApp.

Alexander Wang grew up in New Mexico and has Chinese immigrant scientist parents. He started to think of a new business start up as soon as he got in the ninth grade. In 2015, he went to MIT and chose five graduate computer science courses but he soon dropped out. He soon accepted a job offered by Eric Wu, CEO of Opendoor. Soon after, he enrolled in an AI training program by Sam Altman, now CEO of OpenAI. After a while, he started his own startup known as Scale AI.

Since Scale AI also labeled images, many car companies needed an AI company who could help them with labeling images for their car’s automated systems. They turned to Scale AI. Soon, Scale AI signed up Cruise and Tesla. Scale AI then chooses the contractors and gives them projects. The hiring process at Scale AI is simple. You get hired by applying for the job advertised by the company. The worker is then given written assignments that the model can use to learn how humans think. The writing samples are then given to clients so they can train their AI models. The worker is soon paid for his work.

Image: alexanddeer/Insta

Read next: New Survey Shows What Companies Think of AI Adoption and the Challenges They Are Facing
by Arooj Ahmed via Digital Information World

Cloudflare Rolls Out New Set of AI Tools That Stops Unauthorized Scraping By AI Crawlers

Cloudflare just rolled out a new array of AI tools that provides websites the chance to prevent unauthorized scraping via AI crawlers. Another option is to force these crawlers to pay a fee to get access to data.

The issues of crawling have wreaked havoc on website owners to such an extent that they were forced to pay out of their pocket for the damages done. The issue has been on the rise since the uprising of Generative AI.

The websites will finally get a chance to get the desired value for their content instead of bots stealing material without paying any form of compensation and without taking any consent.

The company confirmed today how it will be launching new tools that make it super easy for website owners, publishers, and creators to take control back over how content is generated by these bots. Similarly, it enables these platforms to block the crawling activity and get compensation wherever they agree to pay for more access. This way, revenue on these platforms is set to increase instead of the age-old tactic of feasting for free on material.

Such AI audit tools also provide users with the chance to witness how content is accessed by others. Remember, it’s a great way to gain insights into how malicious bots used to crash land pages to get access for free. Meanwhile, AI crawlers don’t usually mean harm. They just want a lot of content for free so they can be used to train AI models in bulk quantities.

Cloudflare is a huge name in the industry. It gives rise to more security and better performance regarding optimization for pages. It confirmed that there is no single platform that is dominating in terms of website scraping activities.

It all varies by the kind of content getting scraped at any moment in time.

Meanwhile, generative AI models need huge amounts of data to work and to provide quick and accurate replies to produce images, videos, and music. Today, this AI crawling industry is huge, and big names are involved in terms of giving AI developers what they want and whenever they want. This could be in the form of voice, text, or images.

As per one research company, the web scraping industry is said to hit the 2.45 billion mark by the year 2036.

So many tech giants are making billions by training AI models using work that does not belong to them. This can be used to generate more material in cases where they compete with original material.

In a world like that, this is unacceptable and has gone on for far too long. This is why creators are expressing outrage over the matter. Hence, Cloudflare going one step further to help others in this regard is certainly a positive step welcomed for obvious reasons.

Image: DIW-Aigen

Read next: TikTok Redesigns Subscription Monetization To Eligible Creators In Selected Regions
by Dr. Hura Anwar via Digital Information World

TikTok Redesigns Subscription Monetization To Eligible Creators In Selected Regions

TikTok just confirmed the launch of its new and improved subscription monetization feature for creators. The company says only those creators who come under its eligibility criteria can benefit.

However, only a selected number of regions were outlined including Brazil, Spain, the UK, Indonesia, Germany, South Korea, the US, and then Japan. The company spoke about how it hopes to expand this to several markets in the next few weeks.

The official launch comes after the company spoke about the renaming of LIVE Subscription monetization instead of the old monetization feature. The great part is that it will include non-live creators.

The subscription is the company’s reply to Patreon that allows sharing exclusive content to those paying users. The platform is also competing with other arch-rivals in the industry giving creators more benefits. This includes Instagram and YouTube. Both apps are favorites among the creator ecosystem as they provide growth in earnings through more access and perks.

The expansion will provide more paying users with exclusive benefits that entail better content and perks. What’s interesting is how three separate tiers are up for grabs at different pricing plans. Every single one of them will give exclusive perks for a fee charged every month.

Subscribers can attain exclusive content in the form of more live streaming, videos, and notes that regular users cannot access. They will also get communication channels such as the popular Sub-Space. The latter is the name reserved for a location where creators interact with subscribers.

Subscribers will similarly get special stickers that they can use for LIVE sessions and amongst those include badges. Such things pop up against their names on profiles and inside the comments section.

Creators will also be giving out custom perks that are designed for the community space. One spokesperson for the platform says in an email that customized benefits will be up for grabs and creators can choose what they want from a long list. Common examples provided are Discord roles, BTS videos, shoutouts, and requests for performance.

To access this new Subscription offering, the eligibility criteria must be met. Creators need to be 18 in age and have 10k followers with 100k views on videos. Meanwhile, their creator accounts should have a good standing tag to benefit.

Image: TikTok

Read next: New Survey Shows What Companies Think of AI Adoption and the Challenges They Are Facing
by Dr. Hura Anwar via Digital Information World

New Survey Shows What Companies Think of AI Adoption and the Challenges They Are Facing

Generative AI has become much more than what it was two years ago and is now an important part of many enterprises and businesses. Writer collaborated with Dimensional Research to conduct a survey amongst business enterprises to find out what business leaders think of adoption of AI and what they think of its impact. The survey was done among 501 respondents from five continents who are all related to AI fields in their jobs. The companies these employees belonged to were enterprise organizations having 1000 to 10,000 employees. 65% of the employees surveyed were from IT operations in tech industries.

The results of the survey found that 96% of the companies think that AI is a key enabler, with 82% of the companies expecting rapid growth in their industries because of adoption of AI. Companies say that AI is a key enabler for them because it drives efficiency (72%), saves their time (68%), reduces costs (58%) and improves quality (44%). The top business areas for generative AI are IT operational (65%), customer support (64%), security (54%), marketing (50%) and training (46%).

68% of the companies say that they have more than three teams who are using generative AI. 97% of the companies said in the survey that they expect their new teams to begin using generative AI in the next two years. 31% said that they want their training teams to start using generative AI, 29% said customer support and HR, 27% said marketing and 25% said sales teams should start using generative AI.

95% of the companies said that AI needs enhanced security measures with 94% concerned about data protection with AI applications. Data control and access were the most needed controls for generative AI. Due to security concerns, 78% companies are using in-house or home grown generative AI. 52% of the companies said that their teams are using third-party AI solutions to make in-house AI. 43% are using commercial and 38% are using open-source AI solutions. 88% of the companies say that it takes 6 months or longer to have an in-house AI application operating perfectly.

Only 17% of companies rate AI accuracy as high while 61% claim that they have faced accuracy issues with their in-house AI models. Most companies say that their in-house AI projects are mediocre and sometimes the worst. There's also a shortage in AI experts, with only 10% of the respondents claiming to have high confidence in their AI skills. 63% have moderate confidence while 4% have absolutely no confidence in their AI skills.

The survey also asked the companies which capabilities in AI solutions they highly prefer in commercial settings. The top answers were security controls (69%), followed by data governance (65%) and high quality (57%). This survey shows all the challenges businesses are facing when it comes to AI. There are some factors that need to be crossed when it comes to AI adoption and if a company succeeds to implement all those factors, the generative AI use in the company becomes easy.

Read next: Google AI Overviews Show Up More on Signed In Accounts than on Signed Out Accounts
by Arooj Ahmed via Digital Information World

Data Center Surge: USA Tops Global List with 5,388 Centers, AI Adoption Fuels Growth

In a world fueled by technology, data centers are everywhere. They store, process, and move the information that runs our lives. We rely on them, but managing their impact has become a growing challenge for governments worldwide.

According to the estimates by Cloudscene, there are currently 11,800 operational data centers and the USA has most of them. There are 5,388 data centers in the USA alone as of March 2024. This makes the USA the house of 45% of the total data centers in this world. Germany has the second most data centers (512), followed by China (499) and Canada (366).

France comes at sixth position with 316 data centers listed and Australia at seventh with 307 data centers. The Netherlands (297), Russia (251) and Japan (219) are also the top ten countries with most data centers. These rankings about data centers around the world do not cover the size and capacity of data centers. This could mean that some countries have less data centers but have higher storage capacities. Statistics show that data centers are going to increase with the adoption of AI around the world.

Read next: Experts Believe that Using Blue Light Glasses Cannot Help You With Your Eye Problems But Adopting Some Simple Habits Can
by Arooj Ahmed via Digital Information World