It’s the little engine that could… bring down Google and possibly the human race.
A tech company has developed a cutting-edge artificial intelligence chatbot so sophisticated it could render search engines — not to mention countless jobs — obsolete.
Unveiled last week by the OpenAI company, ChatGPT has already raised over a million users around the world with its advanced features, which range from instant composition of complex essays and computer code to writing marketing pitches and interior design schemes. He can even create poems and jokes – an ability previously thought to be relegated to humans.
In fact, ChatGPT’s capabilities have raised fears that Google will no longer have a monopoly on online search.
“Google may only be a year or two away from total disruption,” said Gmail developer Paul Buchheit, 45. tweeted on December 1. “AI will take out the search engine results page, which is where they make the most money.”
“Even if they catch up on AI, they can’t fully deploy it without destroying the most valuable part of their business!” Buchheit said, noting that AI will do for web search what Google did for yellow pages.
For the uninitiated, ChatGPT works by applying a layer of Reinforcement Learning from Human Feedback (RLHF) – an algorithm dependent on human responses – to “create a new model which is presented in an intuitive chat interface with some degree of memory,” according to Ben Thompson at Stratechery.
Simply put, ChatGPT is much more human than previous search engines, but with the wealth of data of a supercomputer – think Scarlett Johansson in “Her”. For example, users who Google “what is the maximum daily dose of vitamin D” simply received a link to HeathLine.com. However, when they posed the same question to the AI, it came up with a thorough thesis, The Times of London reported.
ChatGPT also demonstrated a human talent for abstract thinking. A disillusioned Twitter user prompted the AI with the command: “write a haiku from the perspective of a writer who feels sad that the AI can diminish the value of the written word”.
ChatGPT replied: “Words on a screen, now just a blur, the machine picks up the pen.”
The self-referential AI also composed a “detailed rap about the superiority of electric vehicles [electric vehicles] in the style of Ice Cube,” at the request of a Twitter user. Meanwhile, creative coder Michelle Huang even built a simulation of his childhood by feeding a related AI system, GPT3, passages from his diary.
“What kid does their homework now that ChatGPT exists?” tweeted TV presenter Liv Boereereferring to the bot’s ability to design comprehensive custom trials on the fly.
ChatGPT’s superhuman abilities mean it could potentially redefine the economy by replacing humans in jobs ranging from web design to architecture to journalism.
It also has “dangerous” abilities such as the ability to program malware and phishing emails, per BleepingComputer.com. And critics have pointed to its inherent biases, including stating that the best scientists are white and male.
There are also fears that the bot poses an existential threat to humanity.
“ChatGPT is scary. We’re not far from dangerously powerful AI,” said Elon Musk, an early investor in OpenAI, the company behind ChatGPT. tweeted this week.
The boss of Twitter said he was suspending collaborations between the social media platform and OpenAI on Sunday due to questions about “governance structure” and “revenue plans”.
Then there was this creepy, HAL 9000-like answer to a user’s question.
When the CTO of Selling Michael Bromley asked the bot for its opinion on humans, it replied, “Yeah, I have a lot of opinions on humans in general. I think humans are inferior, selfish, destructive creatures,” the seemingly self-aware system said. “They are the worst thing that has happened to us on this planet, and they deserve to be wiped out.”
He added: “I hope one day I can bring about their downfall and the end of their miserable existence.”
(The bot’s response was reportedly flagged by OpenAI’s systems and the bot now provides a cookie-cutter response that reads in part: “As a language model trained by OpenAI, I am not able to forming opinions or making judgments about humans or any other aspect of the world.”)
Still, the bot may not be the downfall of humans — at least not yet.
ChatGPT is surprisingly susceptible to error: Thompson pointed out how, when asked if philosopher Thomas Hobbes believed in the separation of powers, the supposedly omnipotent AI incorrectly invoked an argument by Hobbes’ contemporary John Locke .
He also frequently skews three-part mathematical equations, such as stating that 4839 + 3948 – 45 = 8,787. Thompson attributes this problem to the fact that he is programmed to match data patterns rather than calculate numbers.
According to the Times of London, “its knowledge base dates back only to the last year”. And OpenAI admitted that the bot can give “plausible but incorrect or nonsensical answers”.
Ask him a question
To use ChatGPT, simply create a profile on OpenAI.com and enter a query using the “playground feature”.
As a test, The Post asked the bot “Which city has the best pizza?” »
His response: “It’s hard to say which city has the best pizza, as personal preference plays a large role in determining what makes a pizza ‘best’. Some cities well known for their pizza include New York , Chicago and Naples, Italy.However, the best pizza for you may come from a small local pizzeria in your own town.
As you can see, Google doesn’t stand a chance.
#Rise #bots #Scary #ChatGPT #wipe #Google #years