Lo, we have officially entered the realm of Spooky Season, where shadows lengthen and whispers in the darkness speak of wonders and terrors alike. In this eerie time, the world of AI has seen a flurry of developments that chill the soul and raise the hairs on the back of one’s neck. Let us delve into five of the most bone-chilling announcements to emerge from the depths of the AI space.
In a tale fit for the annals of legend, it has come to pass that OpenAI, guided by the enigmatic Sam Altman, has secured a staggering $6.6 billion in investment during its latest funding round. Existing allies like Microsoft and Khosla Ventures have been joined by new companions, including the formidable SoftBank and Nvidia. This influx of wealth has bestowed upon OpenAI a valuation of $157 billion, placing it among the wealthiest entities in all the land. Should Altman’s proposed restructuring plan come to fruition, his equity stake would soar to over $150 billion, propelling him into the hallowed ranks of the top 10 wealthiest individuals on this mortal plane. As the news of this financial triumph reverberates through the ether, OpenAI unveils Canvas, a collaborative feature that stands as a testament to its enduring legacy.
Behold, for Nvidia, the titan of AI hardware, hath ventured into the realm of software with the release of LVNM 1.0, a prodigious open-source large language model that excels in diverse tasks of vision and language. Led by the mighty LVNM-D-72B, boasting a wondrous 72 billion parameters, this new model family is said to rival even the legendary GPT-4o. Yet, Nvidia positions LVNM not as a mere rival to its counterparts, but as a grand platform upon which developers may craft their own creations, be they chatbots or other marvels yet unimagined.
In a twist fitting for this season of shadows and secrets, Google’s Gemini Live has unveiled the ability to converse in nearly four-dozen languages, expanding far beyond the confines of English. This revelation marks a pivotal moment in the quest for direct communication with one’s digital companions, a trend echoed by Microsoft’s unveiling of Copilot Voice, a feature designed to facilitate natural brainstorming on the go. As ChatGPT’s Advanced Voice Mode and Meta’s Natural Voice Interactions usher in a new era of interaction, users find themselves engaged in conversations with their devices, unlocking a realm of possibilities previously unexplored.
Amidst the tumultuous seas of legislative debate, California’s governor, Gavin Newsom, has vetoed the AI safety bill known as SB 1047, a measure aimed at ensuring the safe and secure use of cutting-edge artificial models. In a missive to lawmakers, Newsom argues that the bill’s focus on the largest language models neglects the potential dangers posed by smaller, specialized models that may emerge in the shadows, lurking unseen until it is too late.
Lastly, in a tale reminiscent of ancient enchantments and modern sorcery, a pair of intrepid Harvard students hath wrought a fearsome transformation upon commercially available Meta smart glasses. Through their cunning arts, these spectacles now possess the ability to identify and access the personal information of any who cross their path, a power both wondrous and dreadful. The I-XRAY experiment, as it is known, captures images of unsuspecting strangers, runs them through the arcane algorithms of PimEyes, and discerns their most private details, such as phone numbers and home addresses, from the mists of the digital realm. The implications of such a device are dire indeed, for its wielders walk a perilous path where shadows converge and privacy fades like mist in the dawn. Though the duo have vowed to keep their secrets closely guarded, the specter of their creation looms large, haunting the minds of all who dare to gaze upon its dark visage.