Launch of Claude 3, ChatGPT voice, Elon Musk vs OpenAI — Top AI News of the Week

#1 — Launch of Claude 3

Anthropic, a main rival of OpenAI, has launched Claude 3, which features a suite of models:

• Haiku — a fast model for simple tasks;
• Sonnet — a combination of speed and performance;
• Opus — a power model for complex tasks such as coding or math calculations.

The context window is 200k tokens for all models, less than Google, but more than OpenAI. (Apparently, all models are capable of accepting larger inputs, and that functionality may be made available to certain customers who need additional power.) Opus and Sonnet are currently available through Anthropic’s API, and Haiku is expected to be available soon.

#2 — ChatGPT, say something!

ChatGPT now has a voice: a new (Read Aloud) feature allows chat responses to be spoken in different voices. This feature is not all that new: a similar functionality already existed in the mobile version beginning last fall, which allowed users to speak prompts rather than typing them. But this is OpenAI's branded model for text-to-speech, which will read written answers aloud, and you can also choose a voice from among several offered.

Lately, OpenAI has only been releasing minor updates. Does this mean that a big release is about to happen? Sadly, even verified insiders aren't spilling the beans. Meanwhile, it's been over a year since the release of GPT-4…

#3 — Elon Musk is displeased again

It seems that ChatGPT's voice might be needed soon, as Elon Musk is taking it to court. More accurately, he's not suing the chat – he filed a legal action against OpenAI and CEO Sam Altman. Elon is behaving as usual: on one hand, he's (somewhat unsuccessfully) trying to create AI himself and, on the other hand, he's suing competitors. The billionaire's grievances are as follows:

1. OpenAI was created as a non-profit organization to develop AI to benefit humanity, rather than for profit – protecting humanity from the threats of AI that is owned by corporations. Now, OpenAI collaborates with Microsoft, a mega-corporation that has invested billions in OpenAI.
2. Musk was a founding investor of the non-profit OpenAI, providing more than $40 million dollars (much of its early funding). Now, he wants to know what the money was spent on.
3. Most interestingly: according to the contract between OpenAI and Microsoft, their commercial collaboration should cease when OpenAI has AGI (Artificial General Intelligence) —the form of AI that can perform at or above the level of human intelligence. Musk believes this has already happened, and that GPT-4 is that AGI!

Nobody knows where all of this will lead. This lawsuit will probably be one of the most serious for OpenAI. But, most importantly, the U.S. court will need to provide a clear legal definition of the term Artificial General Intelligence, which we'll be using in the future.

Want to learn more effectively with AI?