Wednesday, October 16, 2024

Google Gemini is Soon Going to Have a New Tool Available to Resize and Reshape Images, Just Like Microsoft Bing Image Creator

Google has developed Gemini go make its AI dreams possible. Just like some other generative AI models, Gemini is good at generating pictures too. Google has recently upgraded Gemini to Imagen 3 and now users can use many tools to have some control over the image that generates. Even though Gemini and Imagen 3 have some of its pros and cons, Gemini is offering a new feature with it too.

Most of the time, the images generated by AI are square shaped. The images created on Gemini are also 1:1 ratio and if someone needs another shape, they have to crop it out (or use Microsoft Bing Image Creator 🙃). But there has been a new setting seen in 15.41.34.29.arm64 beta of the Google app which has some image resizing options. Users will be able to set image sizes to 16:9 ratio too (i.e.: Landscape photos), as spotted by Android Authority.

Another thing is that if you need your subject in the picture to be smaller without changing the ratio of the picture to smaller, you can simply make the subject small and Gemini will add more details in the background. It seems like Google is giving users some free hands to create images on Gemini. Google has also added a tool to refine images on Gemini. Now that the tool for changing the size and shape of the picture is here, this is greatly representing Gemini’s capabilities. But we still have no idea if this update is going to be available at all.

Image: Android Authority

Read next:

ChatGPT Surpasses Bing, Achieves 3.1 Billion Visits in September with 112% Growth

New Study Finds Pedestrians Using Mobiles Can Become a Safety Concern for Drivers as Well as Themselves

• TikTok, Instagram Lead Gen-Z Social Media Engagement, eMarketer Reveals
by Arooj Ahmed via Digital Information World

ChatGPT Surpasses Bing, Achieves 3.1 Billion Visits in September with 112% Growth

Since May 2024, ChatGPT has kept climbing higher smashing its own traffic numbers every month. In September 2024, chatgpt.com saw a mind-blowing 3.1 billion visits, which means it grew by 112% compared to the year before, i.e. according to SimilarWeb's latest data. This follows a bumpier ride in 2023 when the site went through a big rise and fall hitting its high point at 1.9 billion visits in spring before dropping off over the summer months.

The Most Visited Websites in September 2024 – Where is ChatGPT in the Rankings?

Most Visited Websites in September 2024 Traffic
google.com 82.0 billion
youtube.com 28.0 billion
facebook.com 12.3 billion
instagram.com 5.7 billion
whatsapp.com 4.5 billion
x.com 4.3 billion
wikipedia.org 3.8 billion
yahoo.com 3.4 billion
reddit.com 3.4 billion
yahoo.co.jp 3.2 billion
chatgpt.com 3.1 billion
yandex.ru 2.7 billion
amazon.com 2.6 billion
baidu.com 2.4 billion
tiktok.com 2.1 billion
microsoftonline.com 1.8 billion
bing.com 1.8 billion
netflix.com 1.7 billion
linkedin.com 1.7 billion
pornhub.com 1.7 billion

ChatGPT hit the internet in November 2022 and became one of its most popular websites. It started to bounce back from its summer dip by September 2023, but didn't beat its 2023 traffic numbers until May 2024. Since then, the platform has kept growing landing the spot as the 11th most visited website in the world. While it's still behind giants like Google and YouTube, ChatGPT now ranks higher than Amazon's U.S. site cementing its place among the web's top destinations.

In September 2024, ChatGPT reached a new high point by beating Microsoft's Bing in U.S. traffic for the first time. It racked up 442.9 million visits in the U.S. topping Bing's 404.3 million, which showed its growing pull in the AI world. Globally, ChatGPT’s rise has been even more pronounced, aided by a series of product enhancements introduced alongside its switch to the chatgpt.com domain.




Mobile app adoption has also played a key role in ChatGPT’s growth. Although the app launched later, its increasing capabilities, including voice and image recognition, have made it essential for users. The app's user base continues to expand, particularly in markets like the UK, where ChatGPT outpaces competitors in mobile usage.

Even with this success, ChatGPT doesn't pose a threat to search giants like Google yet. It has gained devoted users after offering more features for free. Still, it works on a different scale compared to the biggest mass-market platforms. Yet, its ongoing growth shows a lasting appeal that keeps going strong.

Read next: How to Easily Find the Publish Date of a Website, Webpage or Article. 6 Super Hacks!
by Unknown via Digital Information World

New Study Finds Pedestrians Using Mobiles Can Become a Safety Concern for Drivers as Well as Themselves

A new study by the University of British Columbia finds that distracted walking is equally dangerous as distracted driving. The study was conducted at two busy intersections in Downtown Vancouver and it was revealed that smartphone addiction can affect our safety on the roads and streets. The distracted pedestrians who use their mobile phones on roads walk closer to vehicles than the ones who have no distractions. They are also less likely to change their walking directions because of oncoming traffic and overall tend to walk slower.

This is because distracted pedestrians have less awareness of their surroundings and navigational efficiency gets reduced. On the other hand, non distracted pedestrians have more awareness of their surroundings and have a high level of responsiveness to potential dangers.

This study used Multi-Agent Adversarial Inverse Reinforcement Learning (MA-AIRL) to analyze the interactions between pedestrians and vehicles. According to the results, the pedestrians who were distracted had a 46.5% decrease in their safety levels as compared to non distracted pedestrians. There was a 30.2% less distance between distracted pedestrians and vehicles as compared to non distracted pedestrians and vehicles. Distracted pedestrians also walked 65.7% slower.

The study also provided great insights about driver behavior and how pedestrians on the streets can also lead to potential risks and accidents. It also provided suggestions about urban planning and road safety measures. There should also be warnings for pedestrians who use their mobiles to combat this modern day hazard.

Image: DIW-Aigen

Read next: Frequent Screen Use Impacts Language Skills in Children, Study Reveals
by Arooj Ahmed via Digital Information World

Tuesday, October 15, 2024

Creators on Threads Urged to Post More Often, Engage with Comments to Boost Visibility

Meta's Thread has been in the news because of its rage bait content and having poor content moderation on the platform. Rage bait where a user posts some content on Threads which is guaranteed to anger people and prompt them to comment on the post. Adam Mosseri, Instagram head who also overlooks Threads, said that they are aware of the engagement bait content and are working to have it under control. For starters, Threads is also observing an increase in its users and has surpassed 200 million monthly users.

Business Insider spoke with Emily Dalton Smith, VP of product management at Meta and head of Threads, about Threads and creators who are using the platform. Emily says that more comments on a post means more engagement. Threads say that posts with more engagements or more comments are likely to get more recommendation.

She also added that conversation is like replies to Meta so it is best if creators reply to the comments under the posts so that their posts can get a lot of attention. Emily Dalton Smith also added that the messaging feature isn't coming to Threads soon so users shouldn't expect it. She also told Insider that if a post has an AI tag, it doesn't impact its distribution.

Meta is still working on monetization of Threads and it has rolled out creator monetization to selected users. Threads is also not doing anything on ads right now but it is planning some ways to support creators. Emily also suggested posting more on weekends as weekends drive the highest engagements. Frequent posting can help creators gain higher impressions so it is best to post two to five times a day.

Image: Mohamed Nohassi / Unsplash

Read next:

Overconfidence Alert! This Study Shows Why You’re Probably Wrong About What You Know!

• AI Has Potential to Boost Industry Margins Significantly Over Five Years, Yet Adoption Remains Slow

• Meta Makes Millions From Politically Violent News Events, Shocking New Report Reveals
by Arooj Ahmed via Digital Information World

Overconfidence Alert! This Study Shows Why You’re Probably Wrong About What You Know!

Ever since the evolution of human beings, the ego has predominately driven their intuition and reasoning. As much as the ego may be beneficial to navigate your way through the world, it often blinds humans, too, with overconfidence and feelings of being like a self-existing entity in the world all alone.

When it comes to knowing, people oftentimes believe that they know pretty much everything about anything, but they actually don’t. Social researchers at the University of Ohio have found that people assume naturally that they have all the information about a particular topic even when they don’t. The social scientist named it with a term known as “Illusion of Information Adequacy”.

The study co-author Angus Fletcher said that people follow their instincts and assume to know everything, in fact, they don’t even stop to think if there might be more information out there that might help them to make a much more concrete decision.

To conduct the study in a proper scientific setting and to ensure its reliability and validity, a sample size of 1261 Americans who partook online was taken and divided into three groups. Each group was given the task of reading an article on a fictional school that lacked an adequate water supply.

The first group that read the article started giving reasons for how the school should merge with another school with an ample water supply. Meanwhile, the second group began to reason against the merger after reading the article. The third group gave arguments validating the merging schools and keeping them separate.

Once all the data was collected from the study, it was found that the two groups that only had half of the story had enough conviction to believe that they had more than enough information to make good arguments and decisions. What's really surprising is that the participants of the two groups also assumed that other people would have the same opinion as they did.

However, there were some intriguing results from the findings that the researchers took on a positive note. They noticed that certain participants changed their minds when they were given arguments from the other side of the story after reading their story. The research fits this behavior into the category of not trusting new information or trying to figure out the information to reframe it to fit it into their moral framework.

This study done by Ohio State University falls under the category of Naive Realism, or in layman's terms, a belief in an individual's subjective understanding of being an objective truth.

To further make it clear, naive realism is about how people view the same situation. Still, differently and through information adequacy, it shows that if they have enough information, then they will have the same understanding. However, the major finding of the study is that every human has an innate default mode where they believe with all their heart that they have all the given and necessary information, even when they don’t.

Image: DIW-AIgen

Read next: New Study Shows Parents Prefer AI for Child Healthcare Advice, Raising Concerns
by Ahmed Naeem via Digital Information World

New Passkey Protocols Prioritize Secure Data Exchange Between Platforms, Aiming to Phase Out Traditional Passwords

It would not be wrong to mention that passwords are no longer considered a safe form of authentication. Instead, passkeys have taken over and are now trying to kill the old concept of passwords.

However, passkeys come with a major drawback that has been debated in recent times. This is why the Tech Alliance is working towards overcoming that drawback and ensuring reliable transfer of passkeys from one platform to the next.

The FIDO Alliance which features all leading tech giants mentioned how the industry is making way to move passkeys across different providers. They also shared specifications in draft form for the latest protocol and the format to exchange data.

The Alliance confirmed that safe credential data exchange is a leading priority because it enhances user experience. So far, there’s been no other means for this kind of data exchange between vendors. However, with the recent efforts, it could soon become a reality.

The Alliance adds that users can design passkeys through software taken from tech giants like Apple, Microsoft, and Google. This is in addition to the usual password managers. Similarly, users can share and sync various keys throughout platforms. For now, passkeys are restricted to every firm’s software ecosystem forcing users to develop duplicate keys for single logins.

To address matters like this, passkeys on a single device are used for unlocking access on others using QR Codes. But to help restrict technical barriers, the goal right now seems to be related to making interoperability a reality. Users can use the platform they want to manage their credentials. In the same way, they can switch from one provider to the next without any sort of burden.

For now, draft specifications are focusing more on imports and exports of these passkeys and their encryption. They will be compared with how well they fare against the usual password exchange from one app to the next. The biggest concern here is security and stopping hackers from abuse and theft.

It’s going to take a while for the alliance to confirm the specifications so that they can be adopted commercially. But once that is done, it’s going to be available for users. What they can benefit from is a safe and easy means to change providers, if required.

Image: DIW-Aigen

Read next: Google Turns To Nuclear Energy To Power Its AI Drive
by Dr. Hura Anwar via Digital Information World

Google Turns To Nuclear Energy To Power Its AI Drive

Tech giant Google just entered into a mighty partnership with a leading startup company called Kairos Power. The goal is to use nuclear energy to power the firm’s AI drive.

Google confirmed the news today where it added how it plans to design several nuclear reactors in the US to bring this goal to life. Furthermore, the company revealed how it would include 500MW of nuclear power from small modular reactors by the end of 2030.

Clearly, the plan is ambitious, but the way things are going so far, it does seem like a very high probability that Google will attain success. As the search engine giant says, the first would come into power by 2030 but subsequent ones will come by 2035.

This is the first-ever corporate deal to purchase nuclear power from small module reactors. These are obviously tiny in size when compared to standard ones. The parts are usually designed inside factories instead of being assembled on-site. This assists in bringing construction costs to a new low when compared to full-scale entities.

However, it’s not a very simple deal that can spring into action all of a sudden. Kairos will first require assistance and approval from the American Nuclear Regulatory Commission for design approval. After that, the green signal for construction is required for the plan to move ahead.

So far, the startup got approval and its target date is the year 2027. The firm is already busy designing test units without using nuclear fuel at one of its facilities in NM. Here is where all the parts get assessed including the system and supply chain.

The company failed to reveal the financial aspect of the deal, but according to Google, the deal’s structure will help bring costs down and get energy online sooner. Google says that getting power from several reactors can speed up reactor deployments, which are required to bring costs down and reach the market quickly.

Google’s head for energy and climate sees this approach as the most fruitful one for assisting individuals and the community and building on past efforts. The AI boom means the power demand is at an all-time high, which is why so many tech giants are working towards this goal.

We’ve already witnessed software giant Microsoft forge agreements with companies like Constellation Energy to bring units back online. Furthermore, Amazon purchased a nuclear-powered data center from Talen Energy at the time, too.

Image: DIW-AIgen

Read next: Tech Giants Fuel AI Server Boom, Contributing 30% to Global Server Market Revenue
by Dr. Hura Anwar via Digital Information World