Why Google’s Search Revolution Never Happened

Why Google’s Search Revolution Never Happened

Two years ago, Google had a “revolutionary” chatbot ready. It’s now hurriedly incorporating AI into new items. What took place?

Was Google’s AI progress too slow? Is it the reason Google is rushing to incorporate AI into everything now? Two recent publications present very different portraits of Google both before and after the introduction of ChatGPT.

The revolution in Google search that never occurred. More than two years ago, two Google researchers developed a chatbot that was said to “revolutionize the way people browsed the internet and interacted with computers,” according to The Wall Street Journal.

Executives, however, were said to be risk-averse, concerned that releasing the AI product may harm its $200 billion+ search advertising revenue and its image. And sure enough, Google’s hurried Bard launch significantly damaged its reputation.

What AI principles does Google follow? Google’s targeted action could be caused in part by its AI principles. Google thinks AI software should:

  1. Be of societal benefit.
  2. Be careful not to create or promote intolerance.
  3. Be developed and put through safety tests.
  4. Be responsible to others.
  5. Use privacy-conscious design concepts.
  6. Uphold the highest standards for excellence in science.
  7. Be made available for purposes compatible with these principles.

Another internal worry was a lack of sources. In addition to worries about accuracy and safety, the WSJ notes another significant worry:

“Incorporating tools like LaMDA, which can condense millions of webpages into a single paragraph of text, may further intensify Google’s ongoing disputes with big news organizations and other online publishers by depriving websites of visitors. According to a person familiar with the situation, Google leaders have stressed that the company needs to use generative AI in results in a way that doesn’t annoy website owners, including by giving source links”.

But there were no references to sources when Google demonstrated its new AI search capabilities. And a little outrage was raised as a result.

Next followed Google’s Code Red and OpenAI’s ChatGPT. Ten years ago, Larry Page, the co-founder of Google, issued the following caution: In regard to technology in particular, “incrementalism leads to irrelevance over time because change tends to be revolutionary, not evolutionary.”

Whether you like it or not, ChatGPT is a ground-breaking innovation. Google issued a “code red” shortly after the launch of ChatGPT in late November and requested assistance from Page and co-founder Sergey Brin. This was a part of the year-long effort to give Google Search chatbot features.

Then, on February 6, Google hurried to release Bard, its response to ChatGPT. Microsoft had scheduled to introduce the updated Bing with ChatGPT one day earlier.

Google has now made an effort to clarify that Bard is not a search. While Bard is a distinct product, the AI-powered chatbot capabilities coming to search are built on the same technology.

What if Google AI replaced Google Plus? Bloomberg claims that Google is now “stuffing” more items with generative AI:

“Alumni of Google have been reminded of the last time the business executed an internal directive to inject every significant product with a new idea: the push starting in 2011 to promote the doomed social network Google+. While Google’s proficiency in AI is undeniable, it was never regarded as a leader in social networking, so the analogy isn’t entirely accurate. Nonetheless, the sensation is comparable”.

Google disagreed, stating that testing and improving Bard is a major part of Google’s own operations. Another Googler told Bloomberg:

There is a dangerous mix of unreasonably high expectations and significant uneasiness around any AI-related project.

Why it matters to us. Is Google in a state of panic or moving too slowly? Both statements could be accurate, or the real truth might lie somewhere in the middle, showing that Google is actually following its AI principles. Call it a slow rush; Google can afford to watch and learn from Microsoft and other generative AI players at this time in order to avoid making any (further) costly mistakes.

Leave a Reply

Your email address will not be published. Required fields are marked *