Atomic Media

Microsoft calls deceased NBA player ‘useless’ in AI-written obituary

Written on September 16, 2023 at 6:24 am, by admin

Microsoft has been criticised after publishing an AI-generated obituary for NBA star Brandon Hunter.

The former Boston Celtics and Orlando Magic player passed away suddenly this week, aged 42, after collapsing during a hot yoga class in Orlando, Fla.

Shortly after his passing, fans were shocked to see the father of three described as “useless” in an obituary published on MSN.

The headline read:

• “Brandon Hunter useless at 42.”

Why we care. MSN laid off two dozen editorial staff a few years ago with plans to replace the writers with generative AI, the Guardian reported. This case highlights the importance of not relying solely on AI for generating content due to factual inaccuracies and problematic errors, and the need to ensure that all work produced by AI is supervised by humans. Failure to do so could harm your brand’s reputation as well as negatively impacting your search rankings.

Incomprehensible. While the MSN headline was offensive, the rest of the article was incoherent. It read:

• “Former NBA participant, Brandon Hunter, who beforehand performed for the Boston Celtics and Orlando Magic, has handed away on the age of 42, as introduced by Ohio male’s basketball coach Jeff Boals on Tuesday.”
• “Hunter, initially a extremely regarded high school basketball participant in Cincinnati, achieved vital success as a ahead for the Bobcats.”

Reputational damage. Despite swiftly removing the article from the MSN website, Microsoft was criticized on social media for publishing the offensive content:

What Microsoft is saying. A Microsoft spokesperson told Search Engine Land:

• “The accuracy of the content we publish from our partners is important to us, and we continue to enhance our systems to identify and prevent inaccurate information from appearing on our channels. The story in question has been removed.”

However, the company is yet to officially apologize.

Dig deeper. Futurism broke the news in Microsoft Publishes Garbled AI Article Calling Tragically Deceased NBA Player “Useless”.

Other brands stumbles with AI. We’ve previously reported on a number of brands that have published articles with errors, all of which were lacking in E-E-A-T in different ways:

• Men’s Journal published an AI-generated article, What All Men Should Know About Low Testosterone, that contained bad advice and information.
• BuzzFeed published 44 terrible “AI-assisted” articles.
• Gizmodo published an article on “Star Wars” with numerous factual errors.
• Red Ventures-owned properties (including CNET, BankRate and CreditCards.com) have also leaned heavily into AI-generated content.

As a reminder, Google doesn’t care who – or what – writes your content, as long as that content is helpful and not created to manipulate search results.

The post Microsoft calls deceased NBA player ‘useless’ in AI-written obituary appeared first on Search Engine Land.

Courtesy of Search Engine Land: News & Info About SEO, PPC, SEM, Search Engines & Search Marketing

Google September 2023 helpful content system update rolling out

Written on September 15, 2023 at 3:22 am, by admin

Google has begun to push out an update to its helpful content system, the last helpful content system update was the December 2022 helpful content update. The update started on September 14 and should take about two weeks to fully roll out.

What is new. Google said, “The September 2023 helpful content update is rolling out with an improved classifier. It will take about two weeks to complete. We’ll update our ranking release history page when the rollout is complete.”

Google added new details about helpful content, specifically:

• Added new guidance about hosting third-party content and more explanation on what to do after a helpful content system update (perhaps you don’t need to do anything, or perhaps self-assess your content).
• Added new points about removing content or changing dates to the help page on how to create helpful, reliable people-first content.
• Google also removed the words “written by people” and just wrote “helpful content created for people in search results.” I suspect this is to say AI-generated content is fine when it is helpful. Here is a screenshot of the before and after.

In May, Google told us a helpful content system update would be coming this year. Google said this update would enable the helpful content system to “more deeply understands content created from a personal or expert point of view.” “We’re also improving how we rank results in Search overall, with a greater focus on content with unique expertise and experience,” Google also said a few months ago.

Google said this is not today’s update. Google wrote, “This work is still continuing and is not part of this particular update. We’ll share more about our work in this area in the future.”

What to do if you are hit. Google has provided a list of questions you can ask yourself about your content. Read through those questions as we posted over here, and in an unbiased manner, ask yourself if your content is in sync with this update.

Please note if this update has hit you, it can take several months to recover if you do everything right and make changes to your content over time.

More on the helpful content update. Google’s helpful content update specifically targets “content that seems to have been primarily created for ranking well in search engines rather than to help or inform people.”

This algorithm update aims to help searchers find “high-quality content,” Google told us. Google wants to reward better and more useful content that was written for humans and to help users.

Searchers get frustrated when they land on unhelpful webpages that rank well in search because they were written for the purpose of ranking in search engines. This is the type of content you might call “search engine-first content” or “SEO content.”

Google’s helpful content algorithm aims to downgrade those types of websites while promoting more helpful websites, designed for humans, above search engines.

Google said this is an “ongoing effort to reduce low-quality content and make it easier to find content that feels authentic and useful in search.”

Why we care. If you notice any ranking and visibility changes in Google search over the next two weeks or so, especially if those were big changes, you can likely attribute it to this update. Read Google’s advice, make the necessary changes, and hope for a recovery in the upcoming months.

We hope you all will see a positive trend with your ranking and visibility in Google Search from this update.

The post Google September 2023 helpful content system update rolling out appeared first on Search Engine Land.

Courtesy of Search Engine Land: News & Info About SEO, PPC, SEM, Search Engines & Search Marketing

Amazon sellers can now use AI to write product titles, descriptions and listing details

Written on September 15, 2023 at 3:22 am, by admin

Amazon is rolling out enhanced AI capabilities to help advertisers create better product listings.

The new technology simplifies how sellers create product descriptions, titles, and listing details by automatically generating content based on brief product descriptions.

Previously, sellers were required to enter many pieces of information to create product listings, however, the process has now been streamlined into just one step.

Why we care. Creating product titles, bullet points and descriptions has historically been a time-consuming task for sellers that demanded substantial effort. AI adoption could help alleviate this workload, simplifying and speeding up the process of listing new products, while also potentially improving product listing content.

How it works. To get started, you need to provide a brief description of your product in a few words or sentences. Amazon will then create content for you – which you can then review.

If you wish, you can make further adjustments to the generated content, or you can submit it directly to the Amazon catalog as it is.

Reviews. Early feedback collected over the last few months shows that most sellers are using the content generated by the AI model for their listings without editing it at all.

---

---

What has Amazon said? Robert Tekiela, vice president of Amazon Selection and Catalog Systems, said in a statement:

• “With our new generative AI models, we can infer, improve, and enrich product knowledge at an unprecedented scale and with dramatic improvement in quality, performance, and efficiency.”
• “Our models learn to infer product information through the diverse sources of information, latent knowledge, and logical reasoning that they learn. For example, they can infer a table is round if specifications list a diameter or infer the collar style of a shirt from its image.”

Deep dive. Read Amazon’s announcement in full for more information.

The post Amazon sellers can now use AI to write product titles, descriptions and listing details appeared first on Search Engine Land.

Courtesy of Search Engine Land: News & Info About SEO, PPC, SEM, Search Engines & Search Marketing

Amazon launches new search functions on mobile to rival Google

Written on September 15, 2023 at 3:22 am, by admin

Amazon is rolling out new features to improve its search capabilities on mobile.

In a bid to rival search engines like Google and Pinterest, the retail giant is launching new search and discovery tools, including:

• Multimodal search (the ability to search using both text and images)
• Expanded AR features
• Find-on-Amazon (a new tool that identifies similar products to those in a photo you share directly with the Amazon app)

Why we care. Although each update may seem inconsequential on its own, together, these new tools can influence how consumers search for products and redirect more searches to Amazon. Consequently, these changes could affect Google’s advertising revenue, as Amazon has been closing in on the dominant position held by Google and Meta in digital ad spending.

Multimodal search. This new tool, which is an improvement on Amazon’s existing visuals search engine, allows consumers to add text as well as images to search for matching products. For example, if you need a replacement part for an appliance, you can snap a photo of the part and include the appliance’s name, such as “Frigidaire.”

Expanded AR efforts. Amazon’s augmented reality tool, initially used for visualizing furniture and decor in your space, now extends to smaller items such as lamps and appliances such as toasters and coffee machines. With this AR feature, you can also rearrange items from one surface to another.

Find-on-Amazon. With this tool, you can search for products using a photo from anywhere. If you see something you like on social media, while browsing the web, reading emails, or chatting online, you can tap the “Share” button and send the image to the Amazon Shopping app. This allows the app to find visually similar products even if you don’t know their name or how to describe them.

More updates. Amazon announced that in addition to the changes mentioned above, it is also rolling out other smaller improvements, including:

• The introduction of sales trend data next to listings (for example, how many items were bought in the past month).
• An improved way to easily search for previously purchased items.

---

---

What has Amazon said? An Amazon spokesperson said in a statement:

• “We are always testing new features and honing existing ones to see what works best to help customers in every phase of their shopping journey.”
• “We know that customer trust is hard to win and easy to lose, so we pay close attention to customer feedback about the shopping experience on Amazon.”
• “We’re always experimenting with ways to make it even easier to shop our store. This means that sometimes things may look different as we test new features, but our mission remains the same: to help you find what you’re looking for.”

Deep dive. Read Amazon’s Search and Shop announcement in full for more information.

The post Amazon launches new search functions on mobile to rival Google appeared first on Search Engine Land.

Courtesy of Search Engine Land: News & Info About SEO, PPC, SEM, Search Engines & Search Marketing

TikTok quietly adds Wikipedia snippets to its search results

Written on September 15, 2023 at 3:22 am, by admin

TikTok now serves Wikipedia snippets in some of its search results.

This is the first time that the platform has offered its users results from the wider web as historically, it exclusively featured its own content in SERPs.

The Wikipedia snippets were first noticed in TikTok SERPs by The Verge, as shown here:

A spokesperson for the social media app confirmed that the new feature has been live for several months, however, a formal announcement was never made.

Why we care. While Google has previously stated that it sees TikTok as a search competitor, the social media app is clearly intensifying its efforts in the search arena.

A recent survey revealed the majority of Gen Z women are favoring TikTok over Google for their search needs. This shift in user preference raises questions about whether Google might find itself facing a more concerning competitor in the search space than previously anticipated.

How it works. Wikipedia snippets are served on some accounts for select SERPs for”

• People
• Places
• Events

The snippets been spotted wedged between relevant videos as users have been scrolling down through in-app search results pages.

By clicking on the links that appear at the bottom of the snippet, users are taken to different sections of the Wikipedia entry.

Deep dive. Read TikTok’s Search and Discover guidelines for more information.

The post TikTok quietly adds Wikipedia snippets to its search results appeared first on Search Engine Land.

Courtesy of Search Engine Land: News & Info About SEO, PPC, SEM, Search Engines & Search Marketing

Why now is the right time to implement SMS marketing by Digital Marketing Depot

Written on September 14, 2023 at 12:22 am, by admin

On average, Americans now check their phones 262 times a day and daily screen time averages over five hours per day. People are using their phones more frequently, for longer periods of time, and are shopping from their phones more than ever before.

With 54% of web traffic coming from mobile, adding text messaging (also known as SMS) to your ecommerce stack can help your team reach more customers faster—while maintaining a seamless customer experience.

This practical SMS marketing guide from Cordial provides best practices and essential tips to help you build an effective SMS marketing strategy that drives new revenue streams and loyalty.

Visit Digital Marketing Depot to download Strengthen Customer Loyalty with SMS Marketing.

The post Why now is the right time to implement SMS marketing appeared first on Search Engine Land.

Courtesy of Search Engine Land: News & Info About SEO, PPC, SEM, Search Engines & Search Marketing

TW-BERT: End-to-end query term weighting and the future of Google Search

Written on September 14, 2023 at 12:22 am, by admin

Search is hard, as Seth Godin wrote in 2005.

I mean, if we think SEO is hard (and it is) imagine if you were trying to build a search engine in a world where:

• The users vary dramatically and change their preferences over time.
• The technology they access search advances every day.
• Competitors nipping at your heels constantly.

On top of that, you’re also dealing with pesky SEOs trying to game your algorithm gain insights into how best to optimize for your visitors.

That’s going to make it a lot harder.

Now imagine if the main technologies you need to lean on to advance came with their own limitations – and, perhaps worse, massive costs.

Well, if you’re one of the writers of the recently published paper, “End-to-End Query Term Weighting” you see this as an opportunity to shine.

What is end-to-end query term weighting?

End-to-end query term weighting refers to a method where the weight of each term in a query is determined as part of the overall model, without relying on manually programmed or traditional term weighting schemes or other independent models.

What does that look like?

Here we see an illustration of one of the key differentiators of the model outlined in the paper (Figure 1, specifically).

On the right side of the standard model (2) we see the same as we do with the proposed model (4), which is the corpus (full set of documents in the index), leading to the documents, leading to the terms. 

This illustrates the actual hierarchy into the system, but you can casually think of it in reverse, from the top down. We have terms. We look for documents with those terms. Those documents are in the corpus of all the documents we know about.

To the lower left (1) in the standard Information Retrieval (IR) architecture, you’ll notice that there is no BERT layer. The query used in their illustration (nike running shoes) enters the system, and the weights are computed independently of the model and passed to it.

In the illustration here, the weights are passing equally among the three words in the query. However, it does not have to be that way. It’s simply a default and good illustration. 

What is important to understand is that the weights are assigned from outside the model and entered it with the query. We’ll cover why this is important momentarily.

If we look at the term-weight version on the right side, you’ll see that the query “nike running shoes” enters BERT (Term Weighting BERT, or TW-BERT, to be specific) which is used to assign the weights that would be best applied to that query.

From there things follow a similar path for both, a scoring function is applied and documents are ranked. But there’s a key final step with the new model, that is really the point of it all, the ranking loss calculation.

This calculation, which I was referring to above, makes the weights being determined within the model so important. To understand this best, let’s take a quick aside to discuss loss functions, which is important to really understand what’s going on here.

What is a loss function?

In machine learning, a loss function is basically a calculation of how wrong a system is with said system trying to learn to get as close to a zero loss as possible.

Let’s take for example a model designed to determine house prices. If you entered in all the stats of your house and it came up with a value of $250,000, but your house sold for $260,000 the difference would be considered the loss (which is an absolute value).

Across a large number of examples, the model is taught to minimize the loss by assigning different weights to the parameters it is given until it gets the best result. A parameter, in this case, may include things like square feet, bedrooms, yard size, proximity to a school, etc.

Now, back to query term weighting

Looking back at the two examples above, what we need to focus on is the presence of a BERT model to provide the weighting to the terms down-funnel of the ranking loss calculation. 

To put it differently, in the traditional models, the weighting of the terms was done independent of the model itself and thus, could not respond to how the overall model performed. It could not learn how to improve in the weightings.

In the proposed system, this changes. The weighting is done from within the model itself and thus, as the model seeks to improve it’s performance and reduce the loss function, it has these extra dials to turn bringing term weighting into the equation. Literally.

ngrams

TW-BERT isn’t designed to operate in terms of words, but rather ngrams.

The authors of the paper illustrate well why they use ngrams instead of words when they point out that in the query “nike running shoes” if you simply weight the words then a page with mentions of the words nike, running and shoes could rank well even if it’s discussing “nike running socks” and “skate shoes”.

Traditional IR methods use query statistics and document statistics, and may surface pages with this or similar issues. Past attempts to address this focused on co-occurrence and ordering.

In this model, the ngrams are weighted as words were in our previous example, so we end up with something like:

On the left we see how the query would be weighted as uni-grams (1-word ngrams) and on the right, bi-grams (2-word ngrams).

The system, because the weighting is built into it, can train on all the permutations to determine the best ngrams and also the appropriate weight for each, as opposed to relying only on statistics like frequency.

Zero shot

An important feature of this model is its performance in zero-short tasks. The authors tested in on:

• MS MARCO dataset – Microsoft dataset for document and passage ranking
• TREC-COVID dataset – COVID articles and studies
• Robust04 – News articles
• Common Core – Educational articles and blog posts

They only had a small number of evaluation queries and used none for fine-tuning, making this a zero-shot test in that the model was not trained to rank documents on these domains specifically. The results were:

It outperformed in most tasks and performed best on shorter queries (1 to 10 words).

And it’s plug-and-play!

OK, that might be over-simplifying, but the authors write:

“Aligning TW-BERT with search engine scorers minimizes the changes needed to integrate it into existing production applications, whereas existing deep learning based search methods would require further infrastructure optimization and hardware requirements. The learned weights can be easily utilized by standard lexical retrievers and by other retrieval techniques such as query expansion.”

Because TW-BERT is designed to integrate into the current system, integration is far simpler and cheaper than other options.

What this all means for you

With machine learning models, it’s difficult to predict example what you as an SEO can do about it (apart from visible deployments like Bard or ChatGPT).

A permutation of this model will undoubtedly be deployed due to its improvements and ease of deployment (assuming the statements are accurate).

That said, this is a quality-of-life improvement at Google, that will improve rankings and zero-shot results with a low cost.

All we can really rely on is that if implemented, better results will more reliably surface. And that’s good news for SEO professionals.

The post TW-BERT: End-to-end query term weighting and the future of Google Search appeared first on Search Engine Land.

Courtesy of Search Engine Land: News & Info About SEO, PPC, SEM, Search Engines & Search Marketing

Actionable. Digital. Free. Get your SMX Next sneak peek here.

Written on September 14, 2023 at 12:22 am, by admin

Finish 2023 strong and step into 2024 ready to stay ahead of Google algorithm updates, supercharge your funnel with qualified leads, and leverage generative AI to your advantage: Attend SMX Next, online November 14-15, to learn actionable, brand-safe search marketing tactics for free without leaving your desk.

Programmed by the Search Engine Land editors, this expert-led training program explores the latest, most critical search marketing topics, including…

• Integrating generative AI into your content creation process
• How to recover from Google algorithm ranking updates
• SEO and PPC trends to embrace in 2024
• How searchers are using Google’s SGE and what SEOs need to know

… plus so much more. And you’re the first to hear about our just-announced PPC keynote featuring the one and only Fred Vallaeys!

Stay tuned for more keynote details! Your free pass packs the entire SMX experience:

• Get your specific questions answered by search marketing legends during Overtime, live Q&A following most sessions.
• Participate in engaging discussions with like-minded marketers during Coffee Talk meetups.
• Train at your own pace with instant on-demand access available for 180 days.
• Showcase your commitment to continued learning with a personalized certificate of attendance.

No plane ticket, hotel booking, or expense reports required.

For nearly 20 years, more than 150,000 search marketers from around the world have attended SMX to learn game-changing tactics, connect with like-minded professionals, and further their careers.

Don’t miss your final chance to join us in 2023. Secure your free pass now!

The post Actionable. Digital. Free. Get your SMX Next sneak peek here. appeared first on Search Engine Land.

Courtesy of Search Engine Land: News & Info About SEO, PPC, SEM, Search Engines & Search Marketing

Overcoming the Google SGE challenge: Assessment and recovery strategies by Cynthia Ramsaran

Written on September 14, 2023 at 12:22 am, by admin

Google Search Generative Experience (SGE) is looming over the search industry, but little is understood about its mechanisms and expected impact on our organic traffic.

Join this webinar to understand how to overcome the Google SGE challenge. The first step is assessing how Google SGE goes live and how it will impact your organic traffic. Next, identify how to recover from expected traffic drops. Register now to learn about solutions that will help your organization mitigate the risk of SGE.

Register and attend “Overcoming the Google SGE Challenge: Assessment and Recovery Strategies” presented by AgileSEO.

---

Click here to view more Search Engine Land webinars.

The post Overcoming the Google SGE challenge: Assessment and recovery strategies appeared first on Search Engine Land.

Courtesy of Search Engine Land: News & Info About SEO, PPC, SEM, Search Engines & Search Marketing

Google CEO on SGE and Search evolution: ‘We’ll get it right’

Written on September 12, 2023 at 9:20 pm, by admin

The multibillion-dollar question right now is whether the Search Generative Experience (SGE) will blow up Google’s Search business model. Based on early testing, Google CEO Sundar Pichai is “confident” that won’t happen.

“It’s important to us to connect users with what’s out on the web, and we are working deeply to make sure that continues to work well,” Pichai told Steven Levy in a Q&A published on Wired.

Why we care. Google may be confident that the SGE experiment will continue to send people to websites, but we’re still finding it hard to trust Google at this point. They won’t reveal any real data to us about whether AI answers are driving clicks to websites, click-through rate data, or advertising performance data on what we are calling CHERPs, or Chat Engine Results Pages. And we have no way to track this in Google Search Console, Google Ads or Google Analytics.

Ads vs. organic. Levy pointed out to Pichai that the AI-generated answers are different from a list of links, which could further add confusion about whether an SGE answer is sponsored or organic.

• “Even in a generative experience we would give you a set of sites that support what we are saying. We want to make sure users are consuming those sites. So I don’t think the core part of the experience will change. We will have a space for ads in a way that makes sense for users and particularly on commercial queries,” Pichai said.

Helping people find information. Pichai pointed out that even though the way Google presents information has evolved greatly in 25 years – but ultimately people are searching for information.

• “We are still trying to help people find the best information that exists online. Inherently, people are also looking for commercial information, and ads are very valuable commercial information, because they connect merchants and businesses, small and big, to users. None of that changes just because we are applying AI deeply. When we evolve search with generative AI, we’ll apply the same principles,” Pichai said.

Like desktop to mobile? Pichai pointed out that people asked similar questions as Google shifted its focus from desktop to making everything mobile-first.

• “Our early testing shows that we’ll be able to get it right. It’s core to the company to evolve search while applying the underlying principles. I am confident we’ll be able to get that right through this transition,” Pichai said.

The search community was also extremely worried about Google’s featured snippets stealing traffic. When Bard launched in February, there were no links. And SGE only finally added links to websites Aug. 30 after months of pressure by search marketers and publishers.

Bing burn. Did the new Bing, powered by ChatGPT, make Google “dance.” Pichai essentially dissed their efforts, comparing it to Alexa and Siri. Pichai also claimed:

• “Around the end of last year, my thoughts were, how can we bring generative AI to search in a way that makes sense for our users? That’s what I’m thinking about, and that’s what will matter in the long run.”

Could that time in late 2022 coincide with OpenAI launching ChatGPT in late November? Generative AI exploded on the scene and supposedly triggered a code red within Google, leading to the return of Google founders Larry Page and Sergey Brin.

Pichai was also asked about the antitrust trial, but had little to say beyond throwing around the word “innovation.”

You can read the full interview here: Sundar Pichai on Google’s AI, Microsoft’s AI, OpenAI, and … Did We Mention AI?

The post Google CEO on SGE and Search evolution: ‘We’ll get it right’ appeared first on Search Engine Land.

Courtesy of Search Engine Land: News & Info About SEO, PPC, SEM, Search Engines & Search Marketing