Tag Archives: chatbots

Artificial Intelligence, Robotics

Is ChatGPT Safe? What You Should Know Before You Start Using AI Chatbots

Leave a comment

Before we are able to perceive the risks and security issues posed by ChatGPT and different, related chatbots, we should perceive how they work at a fundamental degree. ChatGPT is a sort of AI referred to as an LLM, which suggests it was educated on huge quantities of textual content, starting from books to tutorial papers, web content material, and all the pieces in between. It analyzes these sources for patterns, and if you sort a immediate, it creates a response primarily based on statistical chances derived from its coaching knowledge. 

For instance, should you ask, “What is 2+2?” ChatGPT doesn’t know math in the best way most people do. Instead, it depends on the patterns it is seen throughout coaching. Since it has ceaselessly detected equations like, “2+2=4” in its knowledge, it’s extremely prone to reply with “4.” However, as a result of that isn’t all the time the case, it would typically inform you that 2+2 is the same as the state of Utah, or one thing equally, inscrutably flawed (or some other creepy factor).

Far from being able to human reasoning, ChatGPT and related chatbots are merely writing one phrase at a time, every primarily based on the statistical chance of the one which got here earlier than. This normally works effectively sufficient, nevertheless it means LLMs will get issues flawed usually sufficient too, and the AI seemingly will not know the distinction.

There’s in fact much more to how AI works, however hopefully, this overview helps you perceive.

News, Technology

Google Slapped With Lawsuit Over Data Used To Train Its AI

The seemingly sudden explosion of publicly-available chatbots that utilize very capable large language models (LLMs) raised uncomfortable questions about the nature of copyright and how creators can be properly involved in (or, at least, compensated for) the AI-training process. At the heart of the matter are the datasets used to train various AI models, which can include everything from content scraped from random blogs to scientific journals, libraries of published books, social media platforms, and more. Some companies that wield vast quantities of human-generated content like Reddit and Twitter have scrambled to ensure they’re paid for the info.

While big companies fight with lawsuits, there are many people indirectly swept up in the matter who don’t have the resources to individually challenge tech giants, which is where class action lawsuits may come into play. It’s no surprise, then, that Google is facing a proposed class action suit that wants, among other things, for the company to hit pause on providing commercial access to its AI models. 

The legal action comes from Clarkson Law Firm, and one of the attorneys on the case, Tim Giordano, explained the reasoning in a statement to CNN: “Google needs to understand that ‘publicly available’ has never meant free to use for any purpose. Our personal information and our data is our property, and it’s valuable, and nobody has the right to just take it and use it for any purpose.” Alphabet, Google, and DeepMind haven’t commented on the lawsuit at the time of writing.