The launch of Elon Musk’s new “Grok” synthetic intelligence (AI) system could not have made waves all through the machine studying neighborhood or instantly threatened the established order, however it’s definitely drawn the eye of Sam Altman, the CEO of ChatGPT maker OpenAI.
In a publish on the social media app X, previously Twitter, Altman in contrast Grok’s comedic chops to that of a grandpa, saying that it creates jokes much like “your dad’s dad.”
GPT-4? Extra like GPT-Snore!
In terms of humor, GPT-4 is about as humorous as a screendoor on a submarine.
Humor is clearly banned at OpenAI, similar to the various different topics it censors.
That’s why it could not inform a joke if it had a goddamn instruction handbook. It is like…
— Elon Musk (@elonmusk) November 10, 2023
In traditional Musk type, the Tesla/X/SpaceX/Neuralink/Boring Firm CEO apparently couldn’t resist the problem. His response, which he claims was written by Grok, begins off by tapping right into a comedic traditional, rhyming “GPT-4” with the phrase “snore” earlier than dusting off an vintage “display screen door on a submarine” reference.
Nevertheless, in additional trendy trend, Grok’s “comedy” rapidly spirals into what seems to be an offended machine diatribe, remarking that humor is banned at OpenAI and including “that’s why it could not inform a joke if it had a goddamn instruction handbook” earlier than stating that GPT-4 has a “stick to this point up its ass that it will probably style the bark!”
So far as CEO v CEO squabbles go, this one could lack the traditional nuance and savoir faire of the legendary Silicon Valley battles of yesteryear (Invoice Gates vs Steve Jobs, for instance), however what todays’ kerfuffle lacks in comedic weight or grace, it’d maybe make up for basically wierdness.
Within the above video, a grinning Invoice Gates lords over Apple’s MacWorld 1997 occasion in a large display screen above Steve Jobs after Microsoft’s $150 million inventory buy within the firm.
Altman and Musk go means again. Each have been co-founders at OpenAI earlier than the latter left the corporate simply in time to keep away from getting swept up within the rocket-like momentum that is carried it to a two-billion greenback valuation.
Within the wake of OpenAI’s success, which has largely been attributed to the efficacy of its GPT-3 and GPT-4 LLM fashions, Musk joined a refrain of voices calling for a six-month pause in AI development largely prompted by as-yet unfounded fears surrounding the supposed potential for chatbots to trigger the extinction of the human species.
Six months later, almost to the day, Musk and X unveiled a chatbot mannequin that the CEO claims outperforms ChatGPT.
Dubbed “Grok,” Musk’s model of a greater chatbot is an LLM supposedly fine-tuned to generate humorous texts within the vein of “The Hitchhiker’s Information to the Galaxy,” a celebrated science fiction novel written by Douglas Adams.
Adams’ literary work is extensively regarded as foundational items within the pantheon of comedic science fiction and fantasy. His humor has been described by pundits and literary critics as intelligent, witty, and stuffed with each coronary heart and humanity.
And that brings us to GPT-4, OpenAI’s recently-launched “GPTs” feature which permits customers to outline a character for his or her ChatGPT interface, and Musk’s full-throated insistence that Grok is funnier.
Tomorrow, @xAI will launch its first AI to a choose group.
In some necessary respects, it’s the finest that at present exists.
— Elon Musk (@elonmusk) November 3, 2023
It’s at present unclear which mannequin is extra sturdy or succesful. There aren’t any customary, accepted benchmarks for LLMs (or comedy, for that matter).
Whereas OpenAI has published a number of analysis papers detailing ChatGPT’s skills, X hasn’t to this point proffered any such particulars about Grok past claiming that it outscores GPT-3.5 (an outdated mannequin of the LLM powering ChatGPT) on sure metrics.