"For self-interested reasons, these institutions tell stories about themselves that aren't quite true, with the predictable result that people who have any kind of problem with them can correctly and credibly charge them with disingenuousness. Google already had this problem, and Gemini makes it a few degrees worse. In his mea culpa/disciplinary letter to staff about Gemini, Pichai wrote:
Our mission to organize the world's information and make it universally accessible and useful is sacrosanct. We've always sought to give users helpful, accurate,
and unbiased information in our products. That's why people trust them. This has to be our approach for all our products, including our emerging AI products.
Here we have an executive unable to speak honestly in familiar and expected ways. Google's actual mission has long been to deliver value to shareholders by selling advertising, much of it through a search engine, which is obviously and demonstrably biased, not just by the content it crawls and searches through but in the intentional, commercially motivated manner in which Google presents it."
Why even refer to what Google does as "products". Are newspapers simply delivery vehicles for advertising. Newspapers contain a product, the product of peoples' work. It is called journalism. Google is a delivery mechanism for _other peoples'_ products, e.g., their journalistic work, or advertisements. None of this "content" is created by Google. If Google contains a "product" it is source code, the work of Google employees, to the extent it is not borrrowed from open source projects. However people are willing to pay for journalism. They are not willing to pay for Google's source code.
Why not call Google's activities "services". Free services for www users. And paid services for advertisers. That seems far more accurate than "products". Google does not produce products. It avoids that risk. It does not hire journalists and produce content. It is a middleman.
Without other peoples' products, these services have no reason to exist. The middleman needs something to intermediate. For so-called "tech" companies it is the public's access to each others' work via the internet, specifically the web.
Google, and the idea of the so-called "tech" company in general, is 100% depenedent on other peoples' work. This is an obvious truth that been observable for decades but only now is it starting receive appropriate attention, thanks to "AI".
Calling its relationship with the public as one of "trust" is far-fetched. Google assumes "trust" much like it assumes "consent".
History has shown constant, non-transparent, manual tweaking going on behind the scenes for the web search engine. Of course it is biased. This tweaking is essential, so much that that the notion of "AI" seems at odds with what Google actually does for revenue, delivering search as a "business". Something always needs manual "fixing". But the notion of AI that "users" subscribe to seems to be one where no human involvement is necessary, where decisions are made based on objective criteria, free from inherent bias. This quest for "AI" may therefore expose the inherent bias of a web search engine and awaken people to the fact that Google due to its commercial nature can never deliver unbiased search results (or "answers").
The original paper annoucing Google promised a transparent search engine in the academic realm. It was an acknowledgment of the fundamental problem with web search engines and commercial bias. This idea was abandoned. Google instead opted to exploit the problem. Subtly, at first. Today, much less so.
Our mission to organize the world's information and make it universally accessible and useful is sacrosanct. We've always sought to give users helpful, accurate, and unbiased information in our products. That's why people trust them. This has to be our approach for all our products, including our emerging AI products.
Here we have an executive unable to speak honestly in familiar and expected ways. Google's actual mission has long been to deliver value to shareholders by selling advertising, much of it through a search engine, which is obviously and demonstrably biased, not just by the content it crawls and searches through but in the intentional, commercially motivated manner in which Google presents it."
Why even refer to what Google does as "products". Are newspapers simply delivery vehicles for advertising. Newspapers contain a product, the product of peoples' work. It is called journalism. Google is a delivery mechanism for _other peoples'_ products, e.g., their journalistic work, or advertisements. None of this "content" is created by Google. If Google contains a "product" it is source code, the work of Google employees, to the extent it is not borrrowed from open source projects. However people are willing to pay for journalism. They are not willing to pay for Google's source code.
Why not call Google's activities "services". Free services for www users. And paid services for advertisers. That seems far more accurate than "products". Google does not produce products. It avoids that risk. It does not hire journalists and produce content. It is a middleman.
Without other peoples' products, these services have no reason to exist. The middleman needs something to intermediate. For so-called "tech" companies it is the public's access to each others' work via the internet, specifically the web.
Google, and the idea of the so-called "tech" company in general, is 100% depenedent on other peoples' work. This is an obvious truth that been observable for decades but only now is it starting receive appropriate attention, thanks to "AI".
Calling its relationship with the public as one of "trust" is far-fetched. Google assumes "trust" much like it assumes "consent".
History has shown constant, non-transparent, manual tweaking going on behind the scenes for the web search engine. Of course it is biased. This tweaking is essential, so much that that the notion of "AI" seems at odds with what Google actually does for revenue, delivering search as a "business". Something always needs manual "fixing". But the notion of AI that "users" subscribe to seems to be one where no human involvement is necessary, where decisions are made based on objective criteria, free from inherent bias. This quest for "AI" may therefore expose the inherent bias of a web search engine and awaken people to the fact that Google due to its commercial nature can never deliver unbiased search results (or "answers").
The original paper annoucing Google promised a transparent search engine in the academic realm. It was an acknowledgment of the fundamental problem with web search engines and commercial bias. This idea was abandoned. Google instead opted to exploit the problem. Subtly, at first. Today, much less so.