The world’s best AI systems can pass exams, write convincingly human essays, and chat so fluently that many find their output indistinguishable from a person's What can’t they do? Solve simple visual logic puzzles
The team behind the logic puzzles aims to provide a better benchmark for testing the capabilities of AI systems — and to help address a conundrum about large language models such as GPT-4. Tested in one way, they breeze through what once were considered landmark feats of machine intelligence. Tested another way, they seem less impressive, exhibiting glaring blind spots and an inability to reason about abstract concepts.
In the past two to three years, LLMs have blown previous AI systems out of the water in terms of their ability across multiple tasks. They work simply by generating plausible next words when given an input text, based on the statistical correlations between words in billions of online sentences they are trained on. For chatbots built on LLMs, there is an extra element: human trainers have provided extensive feedback to tune how the bots respond.
“There’s very good smart people on all sides of this debate,” says Ullman. The reason for the split, he says, is a lack of conclusive evidence supporting either opinion. “There’s no Geiger counter we can point at something and say ‘beep beep beep — yes, intelligent’,” Ullman adds. Research on how best to test LLMs and what those tests show also has a practical point. If LLMs are going to be applied in real-world domains — from medicine to law — it’s important to understand the limits of their capabilities, Mitchell says. “We have to understand what they can do and where they fail, so that we can know how to use them in a safe manner.
Australia Latest News, Australia Headlines
Similar News:You can also read news stories similar to this one that we have collected from other news sources.
Cleaning Up ChatGPT’s Language Takes Heavy Toll on Human WorkersWorkers in Kenya say they were traumatized by the effort to filter violence and abuse out of ChatGPT. They reviewed thousands of graphic text passages, many containing descriptions of self-harm, child sexual abuse and bestiality.
Read more »
Cleaning Up ChatGPT Takes Heavy Toll on Human WorkersWorkers in Kenya say they were traumatized by the effort to filter violence and abuse out of ChatGPT. They reviewed thousands of graphic text passages, many containing descriptions of self-harm, child sexual abuse and bestiality.
Read more »
AI, ChatGPT and inflation push US stocks toward all-time highsUS stocks have rallied so hard that the market is setting its sights on all-time highs once again
Read more »
ChatGPT's late arrival on Android is one more reason I won't ditch my iPhoneOpenAI announced that the ChatGPT app for Android will be released this week, a couple of months after the iPhone app rolled out.
Read more »
5 ways I use ChatGPT to make money and complete time-consuming tasksInsider tells the global tech, finance, markets, media, healthcare, and strategy stories you want to know.
Read more »
GPT-4: Is the AI behind ChatGPT getting worse?The AI powering ChatGPT may provide completely different answers to the same mathematical problems over time, which is fuelling a debate about whether it is getting worse
Read more »