Google AI Overviews Delivering Crazy Incorrect Results In The USA.

In recent months, Google launched their new AI-powered search tool in the USA, called AI Overviews, with the intention of improving the speed and ease of finding information Online.

The AI tool summarises search results at the top of the page so that users don’t have to scroll through multiple links to get the quickest answers to their questions.

Unfortunately, this latest update has been producing wildly false answers to search queries, leading to widespread circulation of incorrect information, just days after its launch.

The result has been rather embarrassing for Google, with numerous instances of misleading search results and confusing false AI suggestions going viral on Social Media.

Examples:

1. Google’s AI summary confidently stated that former US President Barack Obama is a Muslim (When in fact, he is Christian, with Kenyan heritage).
2. Google’s AI Overviews also stated that there are no names of countries in Africa that start with the letter ‘K’ (…apparently Kenya not a recognised country in Africa?).

NOTE: Both queries have been manually removed by Google for violating the company’s content policies.

Each Google AI search overview summary includes the text disclaimer that acknowledges that “Generative AI is experimental.”

Apparently CNN also recently tested Googles new AI tool with the search: “Data used for Google AI training.” The AI Overview response acknowledged that it was “…unclear if Google prevents copyrighted materials from being included in the online data scraping, to train its AI models”, which also sparked major concern about how AI companies operate.

In another instance, AI Overview instructed a user in USA to literally use glue on their pizza to ensure that the cheese would not slide off (This is NOT advisable with any glue, ever!).

Interesting Article: CEO of Google says it has no solution for its AI providing wildly incorrect information.

Sundar Pichai (Google CEO), downplayed these errors with statements like:

“There are still times it’s going to get it wrong, but I don’t think I would look at that and underestimate how useful it can be at the same time,” he said. “I think that would be the wrong way to think about it.”

“Are we making progress? Yes, we are,” he added. “We have definitely made progress when we look at metrics on factuality year on year. We are all making it better, but it’s not solved.”

Many Google users have been sharing their strange AI search results being generated by the new search tool.

“Maybe it wouldn’t give incorrect information if it didn’t have to censor itself. Didn’t the movie “2010: The Year We Make Contact” teach us that when you instruct a computer to lie, you’re going to get some crazy results.” – Google User Ga Dawg Progress.

Additional articles highlighting the issues
Several news outlets have covered the problems with Google’s new AI Overviews search results. For in-depth analysis and more examples, you can refer to these articles:

• BBC Future on Google AI Algorithm Update
• The Verge on Google Search Problems
• TechCrunch on AI Misinformation

Adverse Impacts on South African Companies:
The ripple effects of Google’s AI search errors are not confined to the USA. South African companies are also feeling the impact in several ways:

1. Loss of Trust:
   Incorrect information erodes trust in search engines, leading users to seek alternative platforms. This can reduce the visibility and credibility of South African businesses that rely on Google for online marketing and exposure.
   
2. Economic Consequences:
   Businesses may suffer financial losses due to outdated or incorrect listings. Potential customers could be misled by these AI Overviews search results, which could lead to missed business opportunities and revenue.
   
3. Legal Risks:
   Misinformation, especially regarding legal or financial advice, could expose companies to legal risks. Businesses providing advice based on bizarre search results could face lawsuits or regulatory scrutiny.
   
4. Health and Safety Concerns:
   Incorrect health information can have serious repercussions. Imagine someone does decide to eat rocks or use glue on their pizza to make the cheese stick better?
   
   We’ve all become very used to finding what we are looking for on Google search consol for as long as we can remember. Is it possible that AI has broken that trust during the beta testing phase of Google AI Overviews in the USA?

5. Reputational Damage:
   Companies found to be associated with incorrect or misleading information might face long-term reputational damage. This can affect customer loyalty and brand perception.
   
   

The latest Google AI search update has had significant knock-on effects in the online industry, delivering incorrect and sometimes dangerous search results.

From historical inaccuracies to critical health misinformation, the impact of these strange AI errors is profound.

While Google’s intention with the AI Overviews introduction was to enhance the search experience, the implementation has highlighted the challenges and risks associated with relying heavily on Artificial Intelligence for distributing information.

Users and businesses should continue to verify the information being circulated online across multiple sources to ensure best SEO optimisation. There are free reputable fact-checking sites like Snopes and Fact Check.

It is also wise to understand the current limitations of AI-driven search results and avoid sharing misleading information with your target audiences.