When ChatGPT exploded on the tech scene in November 2022, one of the burning questions in the minds of many technologists was how Google — which, up to that point, was known for its prowess in artificial intelligence — could be caught flat-footed when the starter’s gun sounded in the race to bring generative AI to the masses?
Paul Buchheit (creator of Gmail) gave some answers in a recent episode of the Y Combinator Startup Podcast.
Buchheit, a former Google employee, said that as Larry Page and Sergey Brin began to lessen their involvement in Google’s operations, artificial intelligence development began to be affected by other factors. He said in the podcast that “It became more about protecting the search monopoly.”
Rob Enderle added that this slowed down their AI progress. He is president and principal analysts at the Enderle GroupHe told TechNewsWorld that the firm, which provides advisory services in Bend, Ore., had been in a leading position up until then, but lost it after the refocus.
The Tech Industry Surprises
Ross Rubin at Reticle ResearchNew York City-based consumer technology consulting firm. TechNewsWorld reported that many other large tech companies were taken by surprise when it came to ChatGPT’s popularity and its ability to resonate with a broad audience.
He said that because a large part of Google’s revenue is generated by advertising and search, it was more important to them than Apple or Amazon to be able harness generative AI so that they could either protect or expand their search business.
He added, “Alexa was never a revenue generator for Amazon.” In many situations, Alexa is the voice and face of Amazon. Amazon generates a much higher revenue from AWS than it does the retail business.
He added that “Azure and Windows licensing is a much larger revenue stream for Microsoft than Search”.
Buchheit points out that Google is right to be worried about AI caving in its search goldmine. He explained that AI is a disruptive technology, because it allows people to get the correct answer without having to click through a page of ads.
He continued that Google was aware of the tension between profitability, and providing the correct answers since 1998. He said that there is always the temptation to make results worse in order to get people clicking on more ads.AI has the potential to disrupt that.“
The Perils of Good Results
Greg Sterling, founder of Near MediaThe website is a source of news, commentary and analysis.
He claimed that Google’s focus to increase revenue has already hurt search users. He said that Google’s focus on revenue growth could be affecting the experience of search users.
Search Engine Land in July reported, citing an enterprise SEO platform BrightEdge analysis, that AI overviews for Google searches dropped from 11% to just 7% between June and July.
Mark N. Vena said that by prioritizing its search engine, Google may have slowed down its AI advances compared to competitors like OpenAI or Microsoft who are pushing AI boundaries aggressively. SmartTech Research Las Vegas, Nevada
This strategic pivot could allow rivals to surpass Google or catch up in AI research, especially in generative models, and other cutting edge AI technologies. Over time, this could weaken Google’s position in the AI field.
AI Ad Impact Exaggerated?
The risk of reducing ad clicks due to too-good search results may be exaggerated. Michael Hodel is the director of equity analysis for the media and telecommunications sector at Morningstar Research Services Chicago
“The types of searches that generate ad revenue are those where there isn’t a single answer and where advertisers hope to influence which route the user takes — like, ‘what are some cool new shoes I might like?'” He told TechNewsWorld. “The question now is, how will ads surface in a context of generative AI?”
He said that the answer lies in Sundar Pichai, Alphabet CEO, who made a comment during the second-quarter company conference call.[W]Pichai explained that the ads which appear above or beneath AI overviews provide people with valuable opportunities to connect and take action.
Buchheit stated that Google is also worried about how generative AI will affect ad revenue. Buchheit further claimed that Google is also concerned with the way the technology might be viewed by regulators. AI could anger regulators, as it was only a matter time before the technology said something offensive. “I think that they were a little scared of it,” he said.
He explained, “They were extremely risk-averse.”
Slow Start, but a Formidable Player
He said that if Page and Brin still ran the company then this risk aversion could have been overcome. “But it takes someone with that level of credibility to really bet the company or to say, ‘Yeah, we’re going to do this thing, and it’s going to cause a lot of problems.'” He said.
He said that Google would not have launched AI if they had the opportunity. The only reason Google launched AI was because OpenAI released ChatGPT and they felt forced to.
“And that helped them too because OpenAI took out a lot those bullets when it came to saying crazy, offensive things,” he continued.
Vena said that despite a slow start Google has made significant progress in the AI arms races. Google’s vast data resources, combined with its deep expertise in machine intelligence, have enabled it to develop powerful AI technologies such as Bard/Gemini. It has also made significant advances in large language models (like PaLM) using Google’s AI.
“While competitors, such as OpenAI, and Microsoft, initially garnered attention,” he added, “Google recently accelerated AI initiatives, integrating AI into more of its products including search, consumer applications, cloud services, etc.” Google’s renewed focus and strategic research, combined with its acquisitions, positions it as a powerful player in the rapidly changing AI landscape.
