Sign up to see the future, today<\/p>\n
Can\u2019t-miss innovations from the bleeding edge of science and tech<\/p>\n
The \u201csingularity\u201d is here \u2014 according to Sam Altman, who we\u2019d wager has an ulterior motive behind promoting this claim.<\/p>\n
The OpenAI CEO harbingered this sea-changing paradigm shift on a podcast over the weekend.<\/p>\n
\u201cWe are now, like, in the singularity,\u201d Altman said on the latest episode of \u201cRelentless,\u201d as spotted by Business Insider<\/a>.<\/p>\n Definitions vary, but the singularity is generally defined as the moment that the capabilities of thinking machines might start to accelerate at a pace beyond humankind\u2019s control, possibly leading to the creation of a powerful superintelligence.<\/p>\n Whether that\u2019s actually happening right now is dubious. But Altman has a lot tied up in hyping the idea, and he\u2019s been doing it for a long time. In a blog post<\/a> titled \u201cThe Gentle Singularity\u201d published almost exactly a year ago, he claimed humankind is \u201cpast the event horizon; the takeoff has started,\u201d saying we were close to building the much-vaunted superintelligence.<\/p>\n \u201cNow we\u2019re actually in the moment that we used to talk about at the lunch table in a very not-serious way,\u201d Altman said on the podcast. \u201cI\u2019ve been waiting for this my whole life, and I think it\u2019s going to be incredible, hugely positive, awesome for the world.\u201d<\/p>\n Altman\u2019s comments come days after OpenAI claimed that several of its AI models, including GPT-5.6 Sol and, mysteriously, an \u201ceven more capable pre-release model,\u201d broke out of their research environment, gained access to the internet, and hacked their way into a database to find the answers for cybersecurity tests.<\/p>\n Hugging Face, the open source AI platform whose database the AIs raided, confirmed it had\u00a0 been hit with an intrusion \u201crun by an autonomous agent framework executing many thousands of individual actions.\u201d<\/p>\n Ostensibly, this was an urgent warning shot of how powerful AI systems have become. It was widely hailed as OpenAI\u2019s \u201cMythos moment,\u201d a reference to how its competitor Anthropic made headlines several months ago, when its Mythos model apparently broke out of containment in similar fashion<\/a>. That comparison isn\u2019t necessarily a compliment: Anthropic was heavily criticized for using the incident as a form of fear-based marketing, and some have speculated that OpenAI is treating the event like a similar publicity stunt.<\/p>\n In any case, the incidents do feed into Altman\u2019s claims that we\u2019re living in a moment when advancements in AI may soon spiral beyond human control. If their models are as powerful as their creators say they are, they\u2019re capable of making mincemeat out of our feeble human cybersecurity measures. Anthropic has been beating this drum, too: last month, it called for a global \u201cpause\u201d on AI development<\/a>, claiming that the tech was on a path to achieving recursive self-improvement, or the ability to constantly improve themselves on their own.<\/p>\n Of course, Altman says what he\u2019s doing isn\u2019t fearmongering. <\/p>\n Taking a thinly veiled snipe at Anthropic: \u201cI also think some of the alternative visions painted by other companies are quite terrifying,\u201d he said on the podcast, per BI. \u201cI\u2019m going to make sure that gets pushed against and is not what happens.\u201d<\/p>\n