Artists, drivers, authors, voice actors, and plagiarists are just a few of the occupations that artificial intelligence is already fighting for. Less than a month after using AI to help a man fight a speeding ticket, an AI merely known as “Claude” completed a law and economics exam at George Mason University. Recently discovered AI, developed by the research company Anthropic, passed the exam with a “marginal pass.” The program was funded by accused crypto fraudster Sam Bankman-Fried.
It is being positioned as a challenger to ChatGPT, which is the AI text generator from OpenAI that has swept the Internet. Claude is now in closed beta status in contrast to ChatGPT. Anthropic refers to Claude’s approach as “Constitutional AI,” which is intended to reply to adversarial questions. Other artificial intelligence programs are frequently rendered useless when faced with conflicting issues.
According to economics professor Alex Tabarrok, the exam was assessed anonymously. He also described Claude as “a competitor” and “an improvement” to OpenAI’s GPT. Tabarrok did point out that the response had certain flaws, including the fact that it was “mainly opinion,” and that a stronger response would have included more economic justification.
Enterprise AI app developer Scale discovered that Claude is “more entertaining than ChatGPT”. “Its ability to write coherently about itself, its constraints, and its objectives seem to also allow it to more naturally answer inquiries on other subjects.”
To be fair, not everyone was as impressed with Claude’s achievements. The Financial Times published an article on Tabarrok’s conclusions and stated, “To be honest, this seems more like Claude merely consumed and puked up a McKinsey study.”
Of course, the fact that Claude only managed to receive a “marginal pass” shows that there is still a long way to go before artificial intelligence software can successfully replace attorneys. Until recently, people didn’t even believe that AI could replace lawyers until recently. So in the blink of an eye AI will be able to challenge the best lawyers.
References:
https://futurism.com/ai-passing-grade-law-exam
https://i.insider.com/63c9233ceee94d001a78e694?width=2000&format=jpeg&auto=webp
In my opinion, Claude’s achievement of passing a law and economics exam with a “marginal pass” is a notable development in the field of AI, but it is clear that there is still a long way to go before AI can successfully replace attorneys. The exam was reported to have certain flaws, such as being mainly opinion-based and lacking economic justification. It should be noted that this is a closed beta program and the outcome may not be as strong as expected. More research and testing is needed to fully evaluate its capabilities in this field.