Author: R G Singh

As a Digital Marketing Expert, We share our digital experiences.

Categories
Microsoft

Microsoft’s new AI model VASA-1 can make pictures ‘talk’

VASA 1 by Microsoft

Microsoft introduces VASA -1, an AI model that can generate a video from photo. With the help of this tool you can give expressions to the picture in the video.

The software giant has unveiled a new artificial intelligence model that can generate videos of talking human faces by combining an audio clip and a still image.

Microsoft spokesperson said, “VASA-1 can generate realistic videos of people talking, at a resolution of 512×512 pixels and up to 40 frames per second. It paves the way for real-time engagements with lifelike avatars that emulate human conversational behaviors”

What is VASA-1 AI Video Generator?

Visual Affective Skill Audio or VASA-1 is a top end model of Microsoft, which can create the facial expressions of humans in the best possible way. The tool can generate all the kinds of feelings and emotions from photos on people’s faces.

In a nutshell, the tool can create video with different expressions of that person from a simple photograph. It creates videos with the help of face muscles, lips, nose, head tilt, and other factors.

How the new AI model VASA-1 creates videos from photos

As per Microsoft, VASA-1 uses a user provided photo and an audio clip to create a short video. It is claimed to handle artistic photos, singing audios and Non-English speech. Microsoft said, “It can handle arbitrary length audio and stably output seamless talking face videos”.

The AI model is capable of producing precious lip-audio synchronization as well as generating a wide spectrum of expressive facial nuances and natural head motions.

To add realism into the video, the model accepts optional signals as conditions, like eye gaze direction (forward- facing, leftwards, rightwards and upwards), head distance (close-up, extreme close-up) and emotion offsets (neutral, happiness, anger and surprise).

Microsoft claims that this AI tool can create videos like real life. That means with its help you can give expressions like real life. The company has shown VASA-1 as research demonstration. They have made it clear that there are no plans to release this product or release its API.

Microsoft will not release this product as there is high possibility of it been misused. The tool is similar to OpenAI’s Sora. Both the tools generate realistic videos. Sora creates complex videos from backgrounds and artifacts. VASA-1 is on human expression. Both these tools are not yet available in public domain.

Categories
Google Search Console

Google Search Console security update- Improves the management of ownership tokens

Google Search Console has released a security update around user permissions and controls management where you can better manage the ownership tokens. Ownership tokens are used for when people verify your site in Search Console, Merchant Center and other Google products.

The update will make it easier to manage such access and remove those tokens and permissions when they are no longer needed. Google said- “We are rolling out further improvements to Search Console’s users and permission management, incorporating capabilities related to unused ownership tokens management. We have seen cases where these are accidently left behind after owners have moved on”.

Google had earlier in February 2023 has rolled out improvements to the user and permissions management report. The latest changes will improve the accuracy and reflect the actual state of the unused ownership tokens.

How it works?

Here are the steps to take to manage these ownership tokens-

  1. Visit the Users and Permissions interface
  2. Click “Unused ownership tokens”
  3. Choose the tokens you will like to remove and click “Remove”
  4. Click “Verify Removal” to get update for the unused ownership token.

It is always good to make sure that the right people have the access to your data and Search Console properties and more importantly that the wrong people do not have access to it.

Nir Kalush from Google said-“The update rolled out lets you verify the removal of unused verification token so that the removed owners cannot regain access to the property”.

When removing the previous verified owners make sure to remove all verification tokens for that user. Check your tokens and remove those who should not have access.

Categories
Google

Google Photo users will soon embrace the new AI Editing tools

Starting May 15, 2024, the tech giant Google is extending its artificial intelligence AI-powered editing tools, including Magic Editor and Magic Eraser to all its Google Photo users.

This move will help the users to make complex edits in a perceptive way by leveraging generative AI. The Magic Editor will enable the users to make complex photo edits without having professional level editing skills.

The feature uses the seamless mix of AI techniques including generating AI, to help people edit specific parts of the image like repositioning the subject, changing the color of the sky or photo background and filling the gaps in a photo.

“The Google Photo users that meet the minimum device requirements can use Magic Editor on Android and iOS with 10 saves every month. After 10 saves one will need to either upgrade to a Premium Google One plan”, said Google.

Google further added that-“The feature will be available on devices that have at least 3 GB RAM and runs on Android 8.0/iOS 15 or later software versions. It will be available for users of Pixel tablets.

The feature was first introduced October 2021, and was currently limited to all Pixel smart phones users and Google One members on Android and iOS in February 2023.

https://storage.googleapis.com/gweb-uniblog-publish-prod/original_videos/MagicEditor_GooglePhotos_April2024.mp4?_=1
Magic Editor: Reimagine your photos with the help of generative AI.

The Magic Eraser will allow the users to erase distracting objects from their photos like people in the background and power lines and power poles. AI editing features like Photo Unblur to clear up blurry pictures, Portrait light that lets users adjust the position and brightness of light in portrait photos post capture, HDR effect for photos and videos, Color Pop, and new styles to create personalized, shareable collages are being rolled out to the consumers.

Also, features like Cinematic photos, Portrait Blur, Sky suggestions and video effects are being made available. “The features will start rolling to consumers beginning from May 15 and continue throughout the coming months to the devices that meet minimum requirements”, said Google spokesperson.

Categories
Google Technology

Google might start charging for AI powered Search

Google is considering charging for premium features on its generative AI powered search engine.

The tech giant is exploring various options including incorporating AI powered search features to its premium subscription services, that already provide access to its new Gemini AI assistant in Gmail and Docs.        

This move will mark Google putting it for the first time any of its core products behind the paywall, as it seeks to gain ground in fast-moving AI space. Some time ago Google had experimentally launched the snapshot feature of Generative AI with Google Search.

With the help of this feature, AI shows users about the searched topic above search results. Google earns a lot from Google search but after the arrival of ChatGPT, the company faced a threat to its business. It has been almost one and half years since the inception of ChatGPT was launched, Google is considering a major change in its business model.

Faced with ChatGPT threat, Google started piloting an AI-driven search service, Search Generative Experience last year. This provides in-depth responses to queries alongside traditional search results and ads.

However, the high computing costs of AI generative have made Google hesitant to integrate these features into its primary search engine fully and that is why the major tech giant is looking to charge uses for it.

The traditional search engine would remain free of charge with an additional content available for those who pay a fee. Adverts would continue to appear alongside search results even for subscribers. Google spokesperson said- “We are not working on or considering an ad-free search experience. As we have done many times before, we will continue to build new premium capabilities and services to enhance subscriptions offerings across Google.

Though, the precise information of Google’s plan to incorporate AI powered search into its various subscription tiers as well as the anticipated launch date remain unclear.

Categories
Content

CWV is a ranking factor that ironically does not improve search results

The senior webmaster Trends Analyst and Search relations team lead at Google, John Mueller quoted in the Google’s search off the record podcast- “Achieving a perfect score will not make a difference in the search results. He further added- “CWV as a ranking factor is one part of a bigger ranking engine and how it is applied is not something that Google talks about”.

The podcast explains how to reconcile and how CWV (Core Web Vitals) can be a ranking factor while at the same not something that will noticeable help rankings.

Actual user experience of the website is more important than CWV

Rick Viscomi and Lizzi Sassman downplayed getting poor Core Web Vitals scores, stating what really matters are how actual users are experiencing the website in terms of site performance in the real world.

The duo quoted – “They tested the core web vitals scores for Google page experience documentation and got different scores. She also shared that Google’s own documentation on Page experience only scored 45 on a scale of 1-100. Actual website performance of users browsing the website matters more than the scores”

CWV are great for benchmarking performance in terms of optimizing for sales, ad clicks and conversions which is where a good user experience literally pays off.

CWV improvements now visible in search

CWV as per John Mueller is used in the ranking systems. But the incremental improvements in the CWV scores will not be noticeable in search results. He further added that achieving perfect CWV scores will not make difference in search results.

He added-“I think the big issue is also that the site owners sometimes over-fixate on the metrics themselves. And they spend months of time kind of working on this. And they see this as they are doing something for their search rankings. And probably a lot of those increment changes are not really visible search.”

“The details we tend not to go into. We do not go into thresholds or anything like that. Similarly to how we do not talk about how many words on a page you have to have or all of those details which from my point of view are almost secondary”.

The takeaway is that it is good to improve CWV but a perfect CWV will be rewarded with better rankings.

Conceptualizing core web vitals

Here both talk about the importance of speed in the wider and more general sense and then zoom in to more narrow sense of ranking where they talk about factors that actually make a difference. They stated that performance is good in general sense and other factors are good in more narrow sense of rankings.

Rick Viscomi explains how web performance is important- “It is really good for everybody and the rising tide lifts all boats. Check your website. Make it faster. Eat your vegetables.”

Lizzi Harvey advises that better use of time is to focus on the content quality, which is the narrow focus on improving rankings.

She added-“Focusing on content quality and then still having like a terrible article like the words on the page are not good or the design is not good and you made it really fast. It that really going to make an improvement for your users or for search?”

Ways of looking at Core Web Vitals

The podcast naturally turned out and it makes sense to conceptualize Core Web Vitals as a general big picture factor because it helps reconcile how something can be a ranking factor that on its own does not really make a difference in search results.

Categories
Business Google

Embrace personalized shopping experiences with Google testing tools

The tech giant Google is testing new product discovery elements to give its users a more personalized shopping experience.

This includes style recommendations, brand preferences, AI image generation and virtual try-on technology. This would impact how people find products in the app and potentially how the retailers display their products.

“We are testing this new experience with US shoppers using mobile browsers and the Google App”, says Google spokesperson. He further added “Testing these features offer new opportunities to showcase products and engage with high-value consumers more inclined to make purchases via personalized shopping experienced tailored to their specific preferences”.

  1. Style recommendations – Google is adding new gallery of images related to the product search. When users search for specific apparel like shoes, they will come across Style recommendations section. Here they can offer rating with thumbs up or thumbs down or simply swipe right or left to immediately access personalized results.
    How it works- If the shoppers are not satisfied with their personalized results or would prefer to continue browsing, Google offers the option to rate more items and immediately view another set of results. Also, search engine will remember the preferences for future searches. For instance, when searching for sports watches, you will receive personalized style recommendations based on your past preferences and interactions with products.
    How to manage- If you not prefer to see personalized shopping results or have rated the items incorrectly, you can manage the preferences by just tapping the three dots next to the “Get style recommendations” section and explore personalized options in the “About this result” panel.
  2. Brand preferences- US shoppers searching for smart watches, shoes or apparels on their smart phones, desktop of the Google app can now personalize their shopping experience by selecting preferred brands. Once chosen, the options from these brands will appear instantly.
    How to manage- Tap the three dots next to “Popular from your favorite brands” section, and access individualized options in the “About this result” panel.
  3. Image generation- Google is adding an AI image generation tool for shopping that generates visual matches for the product that one is seeking in order to better align the search query.  It is available for all the US users who have opted into search generative experience within the search labs.
    How to manage- One has to just tap “Generate Images” after the search to see photorealistic options matching the preferences. Once you find one you like, click on it to explore shoppable options with great ease.
  4. Virtual try-on- The virtual try-on will show how the selected clothing products looks on diverse set of real models, who range in size from XXS- 4XL.
    How to manage – Click on the try-on icon in shopping results.

Sean Scott, Google’s VP/GM consumer shopping said –

  • “People shop on Google umpteen numbers of times in a day. Thanks to the shopping graph they see billions of products in their results that are constantly being refreshed. Every hour more than 200 billion listings are updated with the latest information including pricing, in-stock availability and shipping details”.
  • “So whether you are looking for a specialty notepad from Japan, a monogrammed handbag from Paris or just a hammer from a local hardware store, the shopping graph can help find it in a few clicks. So many options make it easy for the user to find exactly what they like.”
  • No two shoppers are alike, which is why we are designing the shopping experience on Google so it is tailored to you.”

Overall, these additions will take shopping discovery tools of Google in a new direction, and could facilitate much better matches.

Categories
Twitter

Elon Musk aims to compete with YouTube through a video plan for smart TVs

“Long-form videos will be soon available on smart televisions”, said the billionaire Elon Musk. He is planning to encourage viewers to watch long videos on the bigger screen.

Musk confirmed plans for X, formerly Twitter to enter a streaming space with a dedicated TV app for Amazon and Samsung smart TVs. According to news agency Reuters, X has rolled out an early version of audio and video calling some users last October.

This move by X in October was initialized after Musk said he planned to turn the platform to super app offering services from messaging to peer-to-peer payments.

When the user posted on X whether the long videos could be watched directly on smart TVs, Musk briefly responded – “Coming Soon”. He further posted – “We just want people to be able to watch long videos in comfort of their big screen TV”

As per Fortune magazine the app will look similar to Google’s YouTube. Beyond smart TVs, X is also exploring video games, podcasts and long-form writing. On Friday, X introduced ‘Articles’, a new way to share long-form written content on the platform.

An unidentified source stated that Musk is gearing up to compete with YouTube and challenge Google with this. YouTube videos were most watched in streaming platforms in the US in 2023. X did not respond to Reuters request for further details on the app.

To become a video-first platform, X has been forging partnerships with the likes of former Fox Commentator Tucker Carlson and former CNN anchor Don Lemon. Ever since Elon Musk bought X in 2022, the platform has struggled to retain advertisers because of controversies.

Categories
Business

Google former software engineer became CEO of Chinese startup after stealing AI secrets

In an unprecedented case a former software engineer of Google allegedly mastered an elaborate, multi-year plan and successfully stole hundreds of AI and super- computing secrets from Google Cloud and then became CEO of AI startup in China.

Linwei Ding, also known as Leon is 38 years old was arrested at his residence in Newark Wednesday. He was charged with the four counts of theft of trade secrets in connection with an alleged plan to steal Google AI proprietary information.

Ding, secretly transferred trade secrets of Google and other relevant information to his personal account while affiliating himself with ‘AI companies that has headquartered in China.

The US Attorney General Merrick Garland said- “We allege the defendant stole AI trade secrets from Google while he was secretly working for two Chinese based companies. We will protect the sensitive technologies developed in America from falling into the hands of those who should not have them”.

Google hires Ding to develop software

Google hired Ding in 2019 to help develop software for its super computing data centers. These data centers host Google’s large AI models and machine learning workloads. Because of its work profile, Ding was granted access to the confidential information of Google’s hardware infrastructure, the software platform and AI models and applications they supported.

In May 2022, Ding began stealing information related to building of super- computing data centers by coping information into a personal Google cloud account.

The Deputy Attorney General Lisa Monaco gave a statement – “The defendant stole from Google over 500 confidential files that has AI trade secrets, while working for China- based companies that seek an edge in AI technology.”

Ding became CEO of his Chinese AI startup and seeks financial aid

In October 2022, Ding was named as chief technology officer for AI startup. In May 2023, he conducted investor meeting to raise capital for new company Shanghai Zhisuan Technology Co. and named himself as CEO.

The potential investors were told Ding was the chief technology officer of the company and that he owned 20% of the company’s stock. The company endorsed the development of software platform designed to accelerate machine learning workloads that include training large AI models.

Ding did not inform Google about his connections with these companies and also did not disclose his travel to China for investors meetings. He even applied to China based startup incubation program and traveled to Beijing to present the company at investor conferences last November.

Ding tries to hide his deeds

To hide his tracks Ding permitted other employees of Google to use his company –issued access badge to scan into the entrance to Google office. This made it look like he was working from his office in USA while he was in China raising money for his startup venture based on the stolen technology

He also allegedly copied data from the source files of Google into Apple Notes applications on his Google issued Macbook laptop. He converted Apple Notes to PDF files and uploaded them from Google network to his separate account. This enabled him to evade detection by Google’s data loss prevention systems.

If convicted Ding can face 10 years imprisonment

If convicted, Ding can face a maximum sentence of 10 years and up to a $250,000 fine for each account of theft. The Deputy Attorney General Monaco stated- “The Justice Department will relentlessly pursue and hold accountable those who would siphon disruptive technologies especially AI for unlawful export.”

The case is currently investigated by US commerce department and FBI while Google has not yet responded or commented on this matter.

Categories
Business

Elon Musk sues Sam Altman and OpenAI over breach of contract

Elon Musk has filed a law suit against is one-time best friend Sam Altman and OpenAI for the breach of contract, unfair business practices and breach of fiduciary duty. Musk has accused OpenAI of profiting from AGI technology, violating the company’s original agreement. 

Musk filed the law suit in the San Francisco court. In the law suit, Musk has stated something that he has been saying for few years now, that OpenAI’s recent collaboration with the Microsoft has breached the original commitment of the company to develop public, open-source artificial general intelligence.

The law suit stated that Altman and co-founder Greg Brockman have violated the OpenAI’s funding mission to develop AI safely and for the benefit of humanity. The billionaire owner of X and SpaceX co-founded OpenAI alongside Altman and Brockman back in 2015 but stepped away from the company in 2018.

Musk disagreed with Altman and Brockman’s plan to turn OpenAI from non-profit to for profit-company and mounted an unsuccessful bid to install himself as CEO before stepping down. In the lawsuit Elon Mush is seeking unspecified damages above $ 105,000.

If the lawsuit is upheld by the San Francisco Superior Court, it will fundamentally reshape how OpenAI and its partner Microsoft do business. The law suit filed said –“OpenAI has been transformed into a closed-source de facto subsidiary of the largest technology Company, Microsoft”.

The 35-page law suit is the latest chapter in the fight between the former business partners that has been simmering for years. The suit is the latest twist for a company entangled in the controversy. In November, OpenAI’s board forced out Mr. Altman and said it no longer trusted him to run the company. He was then reinstated in just 5 days after an employee revolt threatened the company’s future.

The Silicon Valley insiders believe that generative AI technology behind ChatGPT is once in a generation technology that could transform the tech industry as web browsers did more than three decades ago. OpenAI has also reacted to the accusations of Elon Musk by releasing blog post which showcases the old emails of Elon Mush that he has sent to OpenAI founders then supporting the idea of turning OpenAI to for-profit firm.

Now the billion dollar question arises as why Elon Musk sued his best friend Altman?

Now what has happened?

Elon Mush has sued OpenAI for breach of contract stating that the company is no abiding to the original mission of creating open source artificial general intelligence. Musk stated that OpenAI is not abiding to its terms of the contract and open source AGI and has entered a deal with Microsoft.

Musk had supported profit firm

OpenAI has issued a blog post showcasing emails between Musk and other co-founders where the former had supported the idea of converting OPenAI to for- Profit Company.

What was Musk’s plan?

Elon Musk highlighted financial crunch in OpenAI and proposed to amalgamate it with Tesla which the other co-founders did not approve of.

Musk left OpenAI

OpenAI shared blog post stating Elon Musk wanted full control in the company and when it was not approved by other co-founders, he left OpenAI in 2018. As per spokesperson of OpenAI-“Musk wanted full control in the company. He wanted majority equity, initial board control and position of CEO. Doing this was against the mission of OpenAI as one person will have the full control”.

Elon Musk and his Twitter deal

Elon Musk always enters in some rift or the other. Before this when Musk has entered a deal to buy Twitter for $ 44 billion but in the end he did not want to finalize it. When the matter went to court, Musk finalized the deal.

Immediately after finalizing the deal, Musk made many changes in Twitter that includes removing of Interim CEO Parag Agarwal and many senior executives. He then made verification badge of Twitter to paid service and changed the name to X. Musk is relentlessly working to earn more revenue from X and thus he launched subscription model. He also started sharing some portion of revenue to the content creators. Musk aims to achieve $30.22 average revenue of X by end of 2028, which was $ 24.83 last year.

Categories
Google

Google’s recent update to enhance search so you see more useful information

Google is amplifying search so the visitors see more useful information. Every day umpteen numbers of people turn to search to find out specific information.

Google ensures that it showcases relevant information that is unique, information and updated to its visitors. As quoted by Elizabeth Tucker, Director Product Management; Google – “We had policies and automated systems to fight against spammers and we work to address evolving methods that help us incorporate policies for low-quality content”.

She further added, “We regularly update these policies and systems to effectively tackle low-quality content so we can continue delivering useful content and connect people with the high-quality websites”.

Google has announced significant changes to improve quality of search-

  1. Improved quality ranking: The search engine giant has made algorithm enhancements to its core ranking systems to ensure its SERPs surface the most relevant and helpful information on the web and minimize unoriginal content in the search results.
  2. New and improved spam policies: Google is updating its spam policies to keep the lowest quality content out of search like expired websites reclaimed as spam repositories by new owners and memoriam spam.

Minimizing low-quality, unoriginal results

In 2022, Google starting tuning its ranking systems to minimize unoriginal, unhelpful content on the search and keep such content at very low levels. The search engine is bringing what it learned from that work into its March 2024 core update.

This update involves refining some of Google’s core ranking systems to help it better know if the web pages are no helpful to the visitors, have poor user experience or feel like they are created not for visitors but for search engines.

This could include sites created primarily to match the very specific search queries. Elizabeth Tucker further quoted- “The update will minimize the amount of low quality content on the search and send more traffic to high-quality and relevant websites”.

Google believes that this recent update and previous efforts will reduce low quality and unoriginal content in the SERPs by 40%.

Keep more spam out of the results

Google has for past many years relied on the cutting-edge spam-fighting systems and spam policies to prevent the lowest quality content from appearing in the search results and that work continues.

It has been making many updates to its spam policies to better address new and evolving abusive practices that lead to low-quality and unoriginal content showing up on search. The major search engine is gearing up to take action on more types of these manipulative behaviors from today.

While its ranking systems keep many types of low quality content from ranking highly on search, these updates allow it to make more targeted action under the spam policies.

Scaled content abuse

Google has had policy against using automation to generate low quality or unoriginal content at scale with an aim of manipulating search rankings. This policy was designed to address content that was generated using automated tools.

Today, scaled content creation is more sophisticated and it is not clear whether it is created using automated tools. To better address these issues, Google has strengthened its policies to focus on the abusive behavior of producing content in automated tools to boost search rankings.

This allows it to take action on more types of content with little or no value created at tool, like that of pages that pretend to be informative and relevant to popular searches but fail to deliver helpful content.

Site reputation abuse

At times websites that has the best quality content, host low-quality content by third parties with an aim of capitalizing on hosting site’s strong and credible image. Such content ranking highly on the search might confuse the visitors who may have total different expectations on the kind of content on a given website.

Google will now consider such third-party content curated mainly for ranking purpose and without close oversight of website owner as spam and this policy will be published two month in advance of implementation on May 5. This will give the site owners time to make any required changes.

Expired domain abuse

Expired domain and purchased and repurposed with the main aim of enhancing search rankings of low quality and unoriginal content. This misleads the visitors thinking the new content is a part of old site, which is not so. “We will consider such expired domains as spam”, quoted Product Management in-charge of Google.

Search helps people with mammoth of information every day, but Google will always have areas for improvement. With updates and policies, the search giant aims to keep low-quality content on search at low levels and show more of relevant and informative content created for people.

Exit mobile version