Sign up to see the future, today<\/p>\n
Can\u2019t-miss innovations from the bleeding edge of science and tech<\/p>\n
As xAI reels from a mass exodus of cofounders<\/a>, remains mired in numerous controversies, and loses ground to competitors like Anthropic, CEO Elon Musk has quietly made a small but significant change to how its flagship chatbot Grok functions.<\/p>\n Starting last week, Grok stopped taking questions that free users asked in posts on Musk\u2019s social media site, X, where the chatbot is integrated. Instead, it now states that \u201cAsk Grok\u201d is only available to Premium and Premium+ subscribers, paid tiers that the site heavily pushes.<\/p>\n It\u2019s another sign of desperation from Musk. For better or worse, \u201cAsk Grok\u201d became one of the most recognizable features of the website after his takeover. Twitter was no longer Twitter, and tweets were no longer tweets, but Grok was everywhere. Replies that read \u201cHey @Grok\u201d and \u201c@Grok is this true?\u201d became a meme and mantra, even beyond the site, no doubt introducing many users to the chatbot.\u00a0<\/p>\n And for Musk, who has an obsession with combatting narratives from mainstream media, the \u201cAsk Grok\u201d feature allowed his \u201cmaximum truth-seeking AI\u201d to weigh in on all kinds of political issues and set the record in a manner aligned with his personal beliefs. (Sometimes this backfired, as evidenced in episodes like when Grok suddenly began spouting racist conspiracy theories about a supposed \u201cwhite genocide\u201d in South Africa, or when users discovered it would overtly praise its creator<\/a>, leading the bot to make claims that he was a greater role model than Jesus Christ and a genius on par with Isaac Newton.)<\/p>\n In short, Musk has decided to limit what had become a cornerstone of the X experience under his tutelage to paid blue checkmarks. It could be an act of belt-tightening at the company, and it could also be a sign of Musk trying to rein in the notoriously uncouth chatbot; when in late December Grok began churning out a flood of nonconsensual AI-generated nudes of real people, including of minors, xAI\u2019s initial response was to temporarily limit image generating features via Ask Grok to paid users only.<\/p>\n The timing of the change is noteworthy. Following xAI\u2019s recent acquisition by Musk\u2019s space company SpaceX, the aggrandized corporation is looking to go public at a staggering valuation of $1.25 trillion, in what would be the largest IPO in history. Ahead of that, Musk seems to be cleaning up xAI\u2019s act. Last week, Musk said he would rebuild xAI<\/a> \u201cfrom the foundations up\u201d after confessing that it \u201cwas not built right first time around.\u201d Cofounders fled or were pushed out of the company in the flurry, with only three remaining after the dust had settled. Behind the scenes, Musk ordered another round of layoffs, according to the Financial Times<\/a>, reportedly because of his frustration at the performance of its AI coding assistant. \u201cGrok is currently behind in coding,\u201d Musk admitted at a conference last week, per Business Insider<\/a>.<\/p>\n Coding remains one of the main battlegrounds for AI companies as the tools have become popular with software engineers. With the long-term profitability of generative AI tools still being a big question mark, selling AI coding assistants to huge enterprise clients offers one of the few surefire ways of making money. Perhaps a freely available \u201cAsk Grok\u201d feature is a casualty of Musk\u2019s renewed focus on the world of software development.<\/p>\n