Elon Musk announced Grok 3.5, a beta update from xAI that generates unique answers for technical questions. Initially, it’s only available to SuperGrok subscribers. Speculations swirl around Musk raising funds for a massive AI supercomputer even as allegations of improper operations emerge. Grok 3.5’s reasoning model sets it apart from internet-dependent AI, but concerns about needed computational resources linger.
Elon Musk is at it again—over on X, the platform formerly known as Twitter, he announced an upcoming update to xAI’s Grok. The new version, dubbed Grok 3.5, is currently in beta and, not surprisingly, will initially be available only to SuperGrok subscribers starting next week. There’s no word yet on when the general public will get access, leaving many in the tech community eagerly awaiting more news.
This latest announcement comes right amid some pretty wild rumors that Musk is looking to raise a staggering amount—tens of billions, to be exact—to construct a colossal AI supercomputer powered by a whopping one million GPUs. Critics have also raised eyebrows over claims that xAI has been using illegal generators to fuel its existing Colossus supercomputer operation. Quite the messy situation.
The big talking point with Grok 3.5, according to Musk, is its knack for providing unique answers specifically to technical questions, spanning complicated topics like rocket engines and electrochemistry, and supposedly doing so with high accuracy. What sets it apart? Well, while other AI platforms tend to scrape the internet for data, Grok 3.5 takes a different route.
Instead of pulling answers from existing online content, Grok 3.5 is designed to generate information from scratch using a “reasoning” model. This might mean its responses are entirely unique rather than regurgitating what’s already available online. Sounds revolutionary, doesn’t it? This model is reminiscent of DeepSeek R1, which also employs a reasoning mechanism—noted for its dynamic response potential, avoiding the issues of plagiarized content.
However, here’s the catch: the need for that kind of computational heft means that xAI is probably on the hunt for more processing power. As we dive deeper into AI innovations, it’s always a balancing act between capability and resources. There’s an expectation that any day now more details about Grok 3.5’s performance will come to light.
With this beta release in mind, it’s crucial to remember that we might see some variability as it stands now. Users will have to bear with the idea that the final version could differ from what we peek at next week. Until we can scrutinize the results, it’s all just hype and speculation in the AI arena.
Elon Musk revealed the upcoming Grok 3.5 from xAI, which promises unique, self-generated answers rather than relying on the vast, often cluttered internet. This beta version, currently for SuperGrok subscribers, hints at a tech shift in how AI can approach complex queries. Yet, concerns linger about the resources needed and the claims made about its capabilities amidst controversies surrounding xAI’s operations. As the news unfolds, the excitement blends with uncertainty, reflecting the rapidly evolving nature of AI technology.
Original Source: www.tomshardware.com