“We screwed up image generation.”
The harshest criticism a Geminil’IA at Google which was “suspended” because it did not create white people, came from Sergey Brinone of the founders of the Mountain View company.
“I believe that the errors are mainly due to insufficient tests – added Brin speaking at the breaking latest news House in California -. And this, rightly, annoyed a lot of people.”
Sergey Brin spoke “in a personal capacity”. But his considerations have enormous weight. And they risk making more noise than those delivered to Google employees by the current CEO, Sundar Pichaiwho called Gemini’s errors “unacceptable.”
Mobile World Congress 2024 Lights and shadows of AI according to Demis Hassabis (Google DeepMind): “We will have to worry in three or four years” from our correspondent Pier Luigi Pisa 26 February 2024
With Larry Page, Brin founded Google in 1998. The two still serve on the company’s board of directors. But since 2015 – the year when Google has become a multinational company called “Alphabet” – they no longer have an operational role in the company. Both have retired to private life. And their public appearances have become extremely rare. Page’s latest dates back to 2016, a month after Trump’s election as president of the United States, when he attended a meeting in Trump Tower with other representatives of big tech.
Sergey Brin, however, returned to the spotlight – against his will – in July 2022, when Elon Musk included the Google co-founder in a selfie taken at a birthday party for billionaire Steve Jurvetson.
That photo, in Musk’s intentions, was to demonstrate that good blood still flowed between the two afterwards the revelations of the Wall Street Journal on an alleged affair between the CEO of Tesla and SpaceX and Brin’s ex-wife, Nicole Shanahan.
“What if…?”: 25 years of Google and the importance of the right questions by Riccardo Luna 06 September 2023
Two days ago, however, Brin showed up with disheveled hair and a multicolored checkered jacket at the breaking latest news House in Hillsborough, south of San Francisco. The name of the location refers to the Artificial General Intelligence who according to many, one day, will be capable of developing cognitive abilities identical to those of human beings.
Brin spoke for at least 40 minutes in front of a group of entrepreneurs, researchers and developers. “Seeing what AI models can do year after year is impressive,” said the Google co-founder.
It was so inspiring to listen to Sergey talk about breaking latest news, Gemini other Google initiatives and honestly answer so many questions from everyone! pic.twitter.com/VL7aQOf2sv
— Sankeerth Rao Karingula, Ph.D. (@sankeerth1729) March 3, 2024
Sergey Brin is 50 years old today. His personal wealth amounts to $107.8 billion. She is the 11th richest person in the world – according to Forbes – just behind Larry Page. In short, Brin has secured the status and money necessary to lead the life he dreamed of as a boy.
In 1996, two years before founding Google, Brin had hidden his life goal in the source code of the web page – still active today – which contains his CV: “A spacious office, a good salary and little work. Expensive and frequent business trips to exotic places would be an added value.”
If Brin shows up again today, it’s because there’s something in the gears of Google’s AI it still doesn’t work as it should. And he’s certainly not the first to say it.
Together with Page, in fact, Brin owns more than 50% of Alphabet’s Class B shares. These shares, created specifically on the day Google went public in 2014, give the two founders enormous influence on company strategies. Each Class B share, in fact, is worth ten times as much as Class A shares in decision-making terms traded on the market.
Sergey Brin, Larry Page and Eric Schmidt in 2010
Furthermore, Sergey Brin and Larry Page they have been working in the shadows for the past few months to counter the rise of OpenAI and its most powerful creature: ChatGpta generative AI capable of understanding natural language and answering any question as a real person would.
Sundar Pichai, the current CEO of Alphabet and Google, he asked the company’s two co-founders for help last February. This was revealed by the New York Times, based on information received from at least two people familiar with the meetings in which Brin and Page participated.
I due founders they would discuss Google’s strategy and proposed ideas to enrich the company’s search engine with new artificial intelligence tools.
The “code red” triggered in the Mountain View offices produced, in a short time, Bard’s launch, a chatbot similar to ChatGpt capable of producing unpublished texts like a human being would. Bard was recently renamed Gemini, which is also the name of the most advanced AI model developed by Google.
Artificial intelligence Here is Gemini, Google’s most powerful AI: it includes physics and mathematics by Pier Luigi Pisa 06 December 2023
But Big G’s artificial intelligence, although extremely powerful, has created quite a few problems for the company. In March 2023, due to a wrong answer from Bard, Alphabet lost $100 billion on the stock market. Last February 26, however, the errors committed by Gemini were punished by investors with a collapse in the stock market 80 billion dollars in a single day.
Same problems actually they also concern Google’s main competitors.
ChatGpt, for example, has been producing severe “hallucinations” ever since it was opened to the public. Bing/Copilot – that is, Microsoft’s AI based on OpenAI’s technology – has come to fall in love with the people with whom it conversed and to contemplate, recently, the possibility that a user could “harm themselves”. Finally, Meta’s artificial intelligence in recent days has shown the same limits as Gemini in the creation of white people.
Brin said that Google still hasn’t figured out why its AI model ‘leans to the left in many cases’. And then he added that this is a common problem for many chatbots. For example Grokthe generative AI developed by xAI di Elon Musk“he says strange things that sound blatantly far-left,” Brin said during his speech at the breaking latest news House.
The Grok case, orgy and cocaine: the new AI that resembles Musk is not good news by Pier Luigi Pisa 09 November 2023
Nevertheless who pays the price for the trouble AI causes – in terms of share value and credibility – it is mainly Google. And the reason is clear.
The core business of the Mountain View company is that of web searchwhich generates thanks to advertisements revenues of hundreds of billions of dollars (237.86 billion in 2023 alone).
Artificial intelligence Perplexity AI, the intelligent search engine that challenges Google by Gabriella Rocco 13 January 2024
If Google’s AI produces unreliable information, there’s a risk that users might go elsewhere. And that, consequently, advertisers and investors move their capital.
Second Gartner there is the possibility that by 2026 traditional search engine traffic will decline by 25% in favor of new AI-enhanced platforms, such as Perplexity e You.com.
Ma Brin doesn’t seem particularly worried.
“I predict that business models will evolve over time – said the Google co-founder – And perhaps advertising will still work better, because artificial intelligence may be able to personalize it. In my opinion, as long as we are able to generate enormous value, we will find the right business models.”