VENTURE BEAT
EndlessAI has officially launched Lloyd, a groundbreaking AI assistant that brings real-time video analysis to smartphones without the need for additional hardware. Available now on the iOS App Store, Lloyd leverages advanced large language models (LLMs) like ChatGPT and Google’s upcoming Gemini to provide users with immediate insights about the world around them. Since launching just several weeks ago, Lloyd has organically amassed tens of thousands of users, and it appears their AI video assistant might represent the next evolution of generative AI.
One of the significant challenges in scaling generative AI has been the computational cost associated with processing video data. Traditional LLMs require immense computing power to ingest and analyze video frames, making real-time applications impractical. EndlessAI is on a mission to “crack the code” by developing proprietary algorithms and patented processes that aim to optimize video data for LLMs efficiently.
“LLMs require very extensive computing power to understand visual data. With a 1080p video running 86,400 frames per hour, seamless, real-time video analysis has been out of reach. EndlessAI empowers existing LLMs, including the ChatGPT, Gemini and Claude families to ingest and understand complex, real-time video with utmost efficiency,” said Roi Ginat, co-founder and CEO of EndlessAI.
Lloyd stands out by enabling smartphones to “see” the real world, not just static images. Users can interact with Lloyd in various innovative ways:
Lloyd’s can also be used for pure entertainment — as the founders are already seeing these use cases. Lloyd can craft personalized bedtime stories; by taking a panoramic video of a child’s room and specifying a hero or theme, “Lloyd will tell them a mesmerizing story where the elements of the room come alive. Pick a new hero or a new theme or a new room every night. The kids want another story? No problem; how about two more!”
Unlike other visual AI assistants that necessitate hardware purchases, Lloyd operates entirely through existing smartphones. This approach eliminates the need for consumers to invest in additional gadgets that may quickly become obsolete.
“Some of our competitors seem to be taking the opposite approach of building their own smartphones or other smart devices to get better AI features. That approach no longer makes sense: why would anyone want to purchase a separate device to get what was meant to be better AI features that Lloyd has now leapfrogged?” said Thomas Pompidou, Co-Founder and Executive Chairman of EndlessAI. “Rather than requiring users to buy new phones or other hardware and be limited to their modest AI capabilities, Lloyd leverages smartphones connected to the Cloud to provide a vastly superior experience.”
EndlessAI emphasizes robust privacy measures in processing live video and audio inputs. The company employs secure login protocols, data encryption in transit and at rest, and strict access controls. Data retention policies ensure that user data is not held beyond the necessary time to fulfill its intended purpose.
“EndlessAI is committed to complying with all applicable laws and regulations related to data protection and privacy,” added Ginat. “Extensive guardrails with regards to content adequacy are also provided by the own guardrails of the LLM operators; OpenAI, Google, Meta, Anthropic, etc. all have processes and training in place to prevent inappropriate and privacy-breaching content.”
As LLMs continue to advance, Lloyd is designed to improve alongside them. The assistant functions on an over-the-top (OTT) basis, optimizing the input data ingested by LLMs. This means that as models like ChatGPT, Gemini, and others evolve, Lloyd’s capabilities will automatically enhance.
“We are a force multiplier for ChatGPT, Gemini, Claude, Llama, etc. and are poised to significantly expand their market opportunity and we make them immediately viable at scale for video analytics use cases. As a result, as LLMs get better, Lloyd gets better. It’s that simple,” Pompidou explained.
EndlessAI plans to introduce new features to Lloyd, including specialized “Powers” like a “Tour Guide” mode that uses precise geolocation for contextual information and an “Action Mode” for sports and fitness advice. The company is also preparing to launch an API for businesses, enabling enterprises to incorporate real-time video analysis into their processes at scale.
“Integrating video at scale in those solutions opens very extensive new opportunities but faces extensive Accuracy, Speed and Cost challenges,” said Ginat. “It will allow enterprises, businesses, AI scale-ups and developers to build and operate highly scalable Generative AI solutions tailored to their unique needs.”
Founded over a decade ago, EndlessAI has been at the forefront of video analysis and communication technology. The leadership team brings together expertise from software development, venture capital, marketing, and AI specialization:
Lloyd’s launch represents a significant step toward making advanced AI capabilities accessible to the general public without the barrier of additional hardware costs. By harnessing the power of existing smartphones and cutting-edge LLMs, EndlessAI aims to redefine how we interact with the digital and physical worlds.
“Ask yourself ‘what if AI could see the real world like people do?’ and then ask yourself ‘why doesn’t it?’ The answer is ‘it can, but it’s just too expensive to make sense’. We have broken through the cost barrier and now enable LLMs to ‘see’ the real world as it is: a real-time, video world,” mused Pompidou.
VentureBeat newsroom and editorial staff were not involved in the creation of this content.
