The Clippy of AI: Why the battle between Google Bard and Microsoft Bing will fizzle out in the end
Imagine that your company has just hired a rising star in the executive suite who is so enticing that a rival company has just employed someone who looks exactly like them. They are surrounded by an incredible buzz right now. It would appear that all parties involved, from the chief executive officer down to the stockholders, are in agreement that this individual is the key to the company's continued success.
Then you find out that the executive is suffering from what is euphemistically referred to as a "hallucination condition" (Opens in a new tab). When they speak, there is a 15–20% probability that they are making things up. This happens each and every time they do so (Opens in a new tab). The individual is referred to as a "bullshit generator" by a Princeton professor (Opens in a new tab). They are utterly unable to differentiate between fact and fiction (Opens in a new tab). Within the next five minutes, they will be presenting a brand new product on stage. Are you going to continue to put them in the spotlight?
This past week, the response was affirmative for both Microsoft and Google. Microsoft held a last-minute surprise event to announce that OpenAI would bring ChatGPT-style search to the Bing search engine and Edge browser. This announcement was spurred on by the success of OpenAI's ChatGPT, an artificial intelligence chatbot that boasts 100 million monthly active users just two months after its launch. Google had revealed its own artificial intelligence search tool, called Bard, the day before, and had unveiled it at an event in Paris the day following. However, Google ran into its own hallucination problem at the event.
Tuesday, the Chief Executive Officer of Microsoft, Satya Nadella, addressed a group of reporters who had been called to the company's campus in Redmond, Washington. Yes, don't you think it's lovely to believe that (Opens in a new tab)? Microsoft, the IT industry's perennially uncool kid, would dearly want for you to believe that Bing — sorry, "the New Bing" — is competing with Google Search in any way, shape, or form.
The pre-response that Google provided when they announced Bard (Opens in a new tab) was dripping with condescension: In a written statement, Google CEO Sundar Pichai stated, "We re-oriented the organization around AI six years ago."
The 'hallucination problem' brought on by Google
Which is an important point to make. Even while it has had years to include AI, Google, the industry leader in search, is still years ahead of its main competitor in ChatGPT, Bard, which is barely even in the beta stage and only has a small group of testers. Despite Pichai's best efforts to channel his inner hipster, the Bard launch was a disorganized and chaotic affair. It would appear that Google was also taken aback by the amount of publicity surrounding the ChatGPT.
What other explanation is there for the terrible Bard blunder that was on full show at the launch — not at the event itself, where occasional demo flubs are inevitable, but in a GIF that was pre-made? A user is depicted asking Bard for information about the James Webb Space Telescope that he may share with his 9-year-old daughter.
The claim that the James Webb Space Telescope took the very first photograph ever taken of an extrasolar planet is not accurate. Bard was having a psychotic break (Opens in a new tab). (UPDATE: A reporter for the Financial Times states that Bard's phrases were technically accurate; nevertheless, this requires a reading of the language that no human would ever use; this is another issue with AI search.)
It is easy to understand why Alphabet, the parent company of Bard, saw a drop of as much as 8% in its share price on the day that Bard was released. Google highlighted the most significant issue with AI search and further claimed that the business is unable to use its massive data warehouse to fact-check its own work.
Given that Google has already experienced a "hallucination problem" with its highlighted snippets(Opens in a new tab) at the top of search results as recently as 2017, the company ought to be more aware of its mistakes. The snippets algorithm appeared to get a special kick out of spreading false information about former United States presidents. Once more, what could possibly go wrong?
SEE ALSO: You should never put all of your faith in a single source: The new guidelines for acquiring any knowledge online.
To put it another way, if you release your AI search tool too soon, you run the risk of playing against yourself. Microsoft was quite fortunate in that there were no glaring mistakes made public during the launch event that they hosted. But if a search that was based on ChatGPT wasn't rife with errors, why is it still in the beta stage when it's clearly not ready for prime time? As a side note, there is a sign-up sheet available if you are interested in performing unpaid AI quality assurance for Bing (Opens in a new tab).
In answer to an inquiry from Wired regarding the hallucination issue with ChatGPT, Sarah Bird, who holds the title of Head of Responsible AI at Microsoft (which is quite revealing!), stated that "there is still more to do there" (Opens in a new tab). The 15 percent hallucination figure originated from a business that is competing against others to develop a ChatGPT fact-checker, so no, it's not a joke (Opens in a new tab). (UPDATE: a New York Times columnist's frantic piece on New Bing(Opens in a new tab) revealed that it was unable of even correctly doing elementary mathematics or providing a list of local kid-friendly activities.)
Bird went on to say that in earlier iterations of the software, users were given the ability to organize an attack on a school; however, this capability is no longer available. Excellent to know! What is the worst that could possible happen next? It is quite unlikely that this hallucinogenic experimental search tool has any additional unexpected consequences hiding in the shadows that could potentially humiliate a major and legally vulnerable technology firm.
Clippy. Zune. New Bing.
Microsoft is no stranger to humiliation; after all, it is the firm that is responsible for one of the most disastrous failures in the annals of software development—Clippy. The paperclip assistant was notorious for doling out unsolicited recommendation after recommendation. ChatGPT is not like Clippy in the sense that we go to it with questions rather than statements.
But due to the fact that it frequently hallucinates its responses — or, more frequently than you'd expect, provides users a humdrum version on "I can't answer that" — ChatGPT-enabled Bing may be compared to Clippy while under the influence of LSD. If a sufficient number of people using "New Bing" for non-essential searches get jumbled results, then that is what it will be known for in the long run.
It makes no difference if a product is improved in the future; what matters is the initial response of the public, which is what might transform it into a punchline. Microsoft ought to be aware of this as well given that it is the company that created the Zune. To release a ChatGPT product before it is genuinely ready for primetime is the same thing as doing anything else.
The phrase "The New Bing" is practically begging to be used as a punchline at this point. Or, are you really prepared to switch from using Google Search and your Chrome browser to using Bing and Edge if Microsoft is successful in their bid to become the leader in AI search, whatever "winning" actually means in this context? It didn't seem likely to me. As a driving force, technological inertia is grossly underappreciated. https://ejtandemonium.com/
In many contexts, ChatGPT is really impressive; for example, real estate agents particularly enjoy using it (Opens in a new tab) for listing writing. However, in other contexts, it inspires apprehension. When you look below the headlines, though, every report concerning the disruptions caused by it seems to be lacking in some way. It's going to cause a tidal wave of plagiarism amongst students! However, it can also inform you when a paper has been written by ChatGPT (Opens in a new tab), thereby mitigating the harm that it poses to itself. It was successful in the bar examination! On the other hand, it only barely qualified for a grade of C +. (Opens in a new tab).
The process of developing a computer program that is functionally equal to the human brain is referred to as "general AI" in the field of artificial intelligence. Another long-held objective for artificial intelligence is to reach what's known as the insect intelligence level, although we've only just started getting there. Instead of, you know, clicking on links yourself, are you going to actually put your faith in ChatGPT to give the results of your search? http://sentrateknikaprima.com/
The solution may very well depend on how severely you, the reader, are affected by the hallucination problem.
Imagine that your company has just hired a rising star in the executive suite who is so enticing that a rival company has just employed someone who looks exactly like them. They are surrounded by an incredible buzz right now. It would appear that all parties involved, from the chief executive officer down to the stockholders, are in agreement that this individual is the key to the company's continued success.
Then you find out that the executive is suffering from what is euphemistically referred to as a "hallucination condition" (Opens in a new tab). When they speak, there is a 15–20% probability that they are making things up. This happens each and every time they do so (Opens in a new tab). The individual is referred to as a "bullshit generator" by a Princeton professor (Opens in a new tab). They are utterly unable to differentiate between fact and fiction (Opens in a new tab). Within the next five minutes, they will be presenting a brand new product on stage. Are you going to continue to put them in the spotlight?
This past week, the response was affirmative for both Microsoft and Google. Microsoft held a last-minute surprise event to announce that OpenAI would bring ChatGPT-style search to the Bing search engine and Edge browser. This announcement was spurred on by the success of OpenAI's ChatGPT, an artificial intelligence chatbot that boasts 100 million monthly active users just two months after its launch. Google had revealed its own artificial intelligence search tool, called Bard, the day before, and had unveiled it at an event in Paris the day following. However, Google ran into its own hallucination problem at the event.
Tuesday, the Chief Executive Officer of Microsoft, Satya Nadella, addressed a group of reporters who had been called to the company's campus in Redmond, Washington. Yes, don't you think it's lovely to believe that (Opens in a new tab)? Microsoft, the IT industry's perennially uncool kid, would dearly want for you to believe that Bing — sorry, "the New Bing" — is competing with Google Search in any way, shape, or form.
The pre-response that Google provided when they announced Bard (Opens in a new tab) was dripping with condescension: In a written statement, Google CEO Sundar Pichai stated, "We re-oriented the organization around AI six years ago."
The 'hallucination problem' brought on by Google
Which is an important point to make. Even while it has had years to include AI, Google, the industry leader in search, is still years ahead of its main competitor in ChatGPT, Bard, which is barely even in the beta stage and only has a small group of testers. Despite Pichai's best efforts to channel his inner hipster, the Bard launch was a disorganized and chaotic affair. It would appear that Google was also taken aback by the amount of publicity surrounding the ChatGPT.
What other explanation is there for the terrible Bard blunder that was on full show at the launch — not at the event itself, where occasional demo flubs are inevitable, but in a GIF that was pre-made? A user is depicted asking Bard for information about the James Webb Space Telescope that he may share with his 9-year-old daughter.
The claim that the James Webb Space Telescope took the very first photograph ever taken of an extrasolar planet is not accurate. Bard was having a psychotic break (Opens in a new tab). (UPDATE: A reporter for the Financial Times states that Bard's phrases were technically accurate; nevertheless, this requires a reading of the language that no human would ever use; this is another issue with AI search.)
It is easy to understand why Alphabet, the parent company of Bard, saw a drop of as much as 8% in its share price on the day that Bard was released. Google highlighted the most significant issue with AI search and further claimed that the business is unable to use its massive data warehouse to fact-check its own work.
Given that Google has already experienced a "hallucination problem" with its highlighted snippets(Opens in a new tab) at the top of search results as recently as 2017, the company ought to be more aware of its mistakes. The snippets algorithm appeared to get a special kick out of spreading false information about former United States presidents. Once more, what could possibly go wrong?
SEE ALSO: You should never put all of your faith in a single source: The new guidelines for acquiring any knowledge online.
To put it another way, if you release your AI search tool too soon, you run the risk of playing against yourself. Microsoft was quite fortunate in that there were no glaring mistakes made public during the launch event that they hosted. But if a search that was based on ChatGPT wasn't rife with errors, why is it still in the beta stage when it's clearly not ready for prime time? As a side note, there is a sign-up sheet available if you are interested in performing unpaid AI quality assurance for Bing (Opens in a new tab).
In answer to an inquiry from Wired regarding the hallucination issue with ChatGPT, Sarah Bird, who holds the title of Head of Responsible AI at Microsoft (which is quite revealing!), stated that "there is still more to do there" (Opens in a new tab). The 15 percent hallucination figure originated from a business that is competing against others to develop a ChatGPT fact-checker, so no, it's not a joke (Opens in a new tab). (UPDATE: a New York Times columnist's frantic piece on New Bing(Opens in a new tab) revealed that it was unable of even correctly doing elementary mathematics or providing a list of local kid-friendly activities.)
Bird went on to say that in earlier iterations of the software, users were given the ability to organize an attack on a school; however, this capability is no longer available. Excellent to know! What is the worst that could possible happen next? It is quite unlikely that this hallucinogenic experimental search tool has any additional unexpected consequences hiding in the shadows that could potentially humiliate a major and legally vulnerable technology firm.
Clippy. Zune. New Bing.
Microsoft is no stranger to humiliation; after all, it is the firm that is responsible for one of the most disastrous failures in the annals of software development—Clippy. The paperclip assistant was notorious for doling out unsolicited recommendation after recommendation. ChatGPT is not like Clippy in the sense that we go to it with questions rather than statements.
But due to the fact that it frequently hallucinates its responses — or, more frequently than you'd expect, provides users a humdrum version on "I can't answer that" — ChatGPT-enabled Bing may be compared to Clippy while under the influence of LSD. If a sufficient number of people using "New Bing" for non-essential searches get jumbled results, then that is what it will be known for in the long run.
It makes no difference if a product is improved in the future; what matters is the initial response of the public, which is what might transform it into a punchline. Microsoft ought to be aware of this as well given that it is the company that created the Zune. To release a ChatGPT product before it is genuinely ready for primetime is the same thing as doing anything else.
The phrase "The New Bing" is practically begging to be used as a punchline at this point. Or, are you really prepared to switch from using Google Search and your Chrome browser to using Bing and Edge if Microsoft is successful in their bid to become the leader in AI search, whatever "winning" actually means in this context? It didn't seem likely to me. As a driving force, technological inertia is grossly underappreciated. https://ejtandemonium.com/
In many contexts, ChatGPT is really impressive; for example, real estate agents particularly enjoy using it (Opens in a new tab) for listing writing. However, in other contexts, it inspires apprehension. When you look below the headlines, though, every report concerning the disruptions caused by it seems to be lacking in some way. It's going to cause a tidal wave of plagiarism amongst students! However, it can also inform you when a paper has been written by ChatGPT (Opens in a new tab), thereby mitigating the harm that it poses to itself. It was successful in the bar examination! On the other hand, it only barely qualified for a grade of C +. (Opens in a new tab).
The process of developing a computer program that is functionally equal to the human brain is referred to as "general AI" in the field of artificial intelligence. Another long-held objective for artificial intelligence is to reach what's known as the insect intelligence level, although we've only just started getting there. Instead of, you know, clicking on links yourself, are you going to actually put your faith in ChatGPT to give the results of your search? http://sentrateknikaprima.com/
The solution may very well depend on how severely you, the reader, are affected by the hallucination problem.