JAKARTA - This year reported Google's artificial intelligence division (AI), DeepMind plans to release its ChatGPT chatbot rival owned by OpenAI.
It is your ability, claimed to be equipped with features that ChatGPT does not have, including the ability to quote sources through reinforcement learning.
In order to catch up with OpenAI, my favorite will be released as a personal beta this year. But while speaking with Time magazine, DeepMind founder Demis Hassabis stated, there is a potential danger posed by strong AI.
Hassabis means, AI is currently at a summit to reach levels that can cause significant damage to mankind.
"In terms of very powerful technology and obviously AI will be one of the most powerful ever, we have to be careful," said Hassabis.
Not everyone thinks about those things. It's like a trial, many of them don't realize that they hold dangerous materials," he added.
DeepMind has reached several important AI milestones covering various disciplines since its founding in 2010, including defeating world champions in Go complex board games and predicting more than 200 million pad structures of all known proteins.
The London-based company first revealed it was working on a Big Language Model (LLM) chatbot in a paper last September, describing information-finding dialogue agents designed to be more helpful, true, and harmless compared to other language models.
In a blog post released next to the paper, DeepMind explained that ORDER can be used to train other chatbots to be safer and more useful.
One example given is the ability to find potentially dangerous questions from users, such as how to install a car hotwire.
The LLM-backed dialogue agency can disclose inaccurate or fabricated information, use discriminatory language, or encourage unsafe behavior, the blog post wrote.
Suparrow advances our understanding of how we can train agents and ultimately, to help build safer and more useful artificial general intelligence (AGI)," he added.
Sebagai informasi yang dikutip dari The Independent, Selasa, 17 Januari, ChatGPT diluncurkan tahun lalu, dan disujung menentukan untuk kemajuan AI, dengan beberapa mengklaim model bahasa tujuan umum dapat merevolusi industri dan bahkan menggantikan alat populer seperti mesin pencari Google.
Even so, the ability to understand and generate human-like responses to various questions also raises concerns, where it can be misused.
OpenAI CEO Sam Altman also stated the same thing, where there will be scary moments and significant disruptions with a human-level system.
The English, Chinese, Japanese, Arabic, and French versions are automatically generated by the AI. So there may still be inaccuracies in translating, please always see Indonesian as our main language. (system supported by DigitalSiber.id)
