Connect with us

Hi, what are you looking for?

DOGE0.070.84%SOL19.370.72%USDC1.000.01%BNB287.900.44%AVAX15.990.06%XLM0.080.37%
USDT1.000%XRP0.392.6%BCH121.000.75%DOT5.710.16%ADA0.320.37%LTC85.290.38%
THE BIZNOB – Global Business & Financial News – A Business Journal – Focus On Business Leaders, Technology – Enterpeneurship – Finance – Economy – Politics & LifestyleTHE BIZNOB – Global Business & Financial News – A Business Journal – Focus On Business Leaders, Technology – Enterpeneurship – Finance – Economy – Politics & Lifestyle

Technology

Technology

Google stops searching for AI overviews after telling us to eat glue and rocks.

image credit: google image credit: google
image credit: google image credit: google

Listen to the article now

Due to the complexity of AI, the search result summary that the internet giant’s AI generates for public testing is difficult.
The internet company announced that it would limit AI Overview responses in cases where they detect “nonsensical” or entertaining inquiries. Google’s cautious implementation of AI demonstrates their prudent approach.
The company admitted that there were indeed instances of peculiar, incorrect, or unhelpful AI summaries that appeared. The AI occasionally provided inaccurate historical information or generated humorous responses to serious inquiries. However, it justified its AI search results.
Liz Reid, a representative from Google, stated that due to the vast size of the internet and the immense number of daily queries, it is inevitable to encounter peculiarities and mistakes. She said that Google has made over a dozen system improvements to boost AI search results, showing its dedication to continual progress.
Google’s choice to terminate its AI Overviews service is the most recent addition to a series of unsuccessful product debuts by technology behemoths vying in the AI revolution.

Facebook's new rebrand logo Meta is seen on smartphone in front of displayed logo of Facebook, Messenger, Instagram, WhatsApp and Oculus in this illustration picture taken October 28, 2021. REUTERS/Dado Ruvic/Illustration/File Photo

Facebook’s new rebrand logo Meta is seen on smartphone in front of displayed logo of Facebook, Messenger, Instagram, WhatsApp and Oculus in this illustration picture taken October 28, 2021. REUTERS/Dado Ruvic/Illustration/File Photo

Microsoft, Meta (the parent company of Facebook), and X have rapidly integrated artificial intelligence (AI) into their flagship products, resulting in different outcomes. Notable instances of embarrassment have occurred due to AI systems displaying affection towards users, image generators distorting the likeness of historical figures, and the spread of summaries regarding events that do not actually exist in real-time. Businesses and governments still view AI as a crucial tool for the future.
At its I/O developer conference in mid-May, Google stated that it will immediately launch AI Overviews to a larger audience after months of preparation. The company discussed its Gemini model and AI ambitions for the entire two-hour presentation at I/O. Search result summaries are generated using the Gemini model, which is part of AI overviews.

It’s possible that Google’s main search service has a bigger problem with AI overviews. AI technology has trouble with “hallucination,” or making up false facts, so it can’t always give useful answers. Every day, billions of people use Google to find information. Tech companies like Google think they can solve these problems, but some have taken back their goods and told users that AI may give them the wrong information.

Google wrote a blog post on Thursday night about many of the ways it is changing AI Overviews search results:
For questions that don’t make sense, it uses “better detection mechanisms.” These methods use strong natural language processing tools to figure out the purpose and structure of a question. This makes criticism and entertainment harder to do.
It is against the rules to use user-generated content in “responses that could offer misleading advice.” While user input is important, it’s not the only thing that AI overviews are built on. The method gives accurate information by using checked data and sources.

Reid said Thursday, “We know that people trust Google Search to give them correct information.” “We hold ourselves to a high standard, as do our users, so we expect and appreciate the feedback and take it seriously.”


Comment Template

You May Also Like

Business

Amazon CEO Andy Jassy emailed employees revised return-to-work guidelines in February 2023. Starting on May 1, 2023, most Amazon employees were required to work...

Technology

Elon Musk, who has warned of a global depopulation problem, had another child with a Neuralink executive. Neuralink director Shivon Zilis had the child...

Business

  Elon Musk helps me understand which laws in America are complex (contract law, possibly copyright law) and which are conventions. If you’re rich...

Technology

LG Electronics of South Korea has teamed up with ChargePoint to deploy more electric vehicle charging stations in the United States, the firms announced...

Notice: The Biznob uses cookies to provide necessary website functionality, improve your experience and analyze our traffic. By using our website, you agree to our Privacy Policy and our Cookie Policy.

Ok