At Lenovo’s Tech World 2024, each Lenovo and Motorola offered groundbreaking synthetic intelligence (AI) improvements, aiming to push the boundaries of hyper-personalization in client expertise. Lenovo debuted “AI Now,” a generative synthetic intelligence (genAI) system constructed on Meta’s Llama 3.1 AI mannequin, which guarantees to make PCs smarter by turning them into customized digital assistants. The standout function? Hyper-personalization with out compromising consumer privateness.
In contrast to most AI techniques that depend on cloud computing, AI Now retains interactions saved domestically on the gadget. Furthermore, the genAI assistant operating on the PC is powered by the system {hardware} and a Native Giant Language Mannequin (Native LLM), a one-of-a-kind innovation that not solely limits the AI’s interplay with the cloud however can ship responses to consumer queries with out an web connection.
Likewise, Motorola (a Lenovo firm) previewed “Moto AI,” a genAI assistant that goals to make smartphones hands-free. Customers can now handle their gadgets with voice instructions, and options like “Catch Me Up” and “Pay Consideration” assist summarize private communications or transcribe conversations for overview. The “Bear in mind This” function additionally permits customers to seize reside moments or on-screen info by means of voice and gesture instructions.
“Our agentic AI-powered techniques retain consumer context and reminiscence, permitting for extra pure, steady interactions with out the necessity to reintroduce info. By compressing a number of historically massive AI fashions into environment friendly 7-billion parameter autonomous techniques, we plan to ship high-performance AI throughout Lenovo and Motorola ecosystems whereas sustaining accessibility,” Yang Yuanqing, CEO of Lenovo Group, instructed me. “Our purpose is to develop a novel non-public information base inside client gadgets that adapts to particular person wants, with out requiring any technical experience to function.”
Each manufacturers are betting on genAI to drive the next-gen of AI-powered client tech. However can these AI developments ship on their guarantees with out compromising consumer privateness?
A New Period of Assistants: Can We Belief AI with Delicate Knowledge?
Lenovo’s AI Now gives hyper-personalized AI options, storing consumer information domestically throughout the consumer’s system {hardware}. The AI can summarize paperwork, transcribe conferences, and even modify gadget settings based mostly on pure language enter—all with out an web connection. The innovation is powered by a novel structure combining the CPU, GPU, and NPU to energy the AI system and consumer interactions domestically, making a Private Information Base (PKB) saved totally on the gadget.
AI Now doesn’t simply carry out duties; it retains context and reminiscence, so customers don’t must reintroduce their preferences. This creates a extra pure, steady interplay. Lenovo asserts that the native LLM structure is a “first within the trade innovation”, that may ship customized client experiences whereas safeguarding information privateness.
“The way forward for AI is customized, proactive, and predictive. AI Now’s our first step in the direction of growing an Agentic AI system for customers—an augmented intelligence that is aware of your preferences, understands your distinctive cognitive behaviors, and autonomously works for you,” Tolga Kurtoglu, CTO of Lenovo, instructed me. “It transcends the static, chatbot-driven use circumstances, delving into your every day exercise and interplay information together with your PC, processing these insights, and remodeling them into actionable outcomes to spice up client productiveness.”
Motorola’s Moto AI additionally harnesses the ability of Native LLM and LAM expertise to streamline on a regular basis duties in your telephone—from setting alarms to producing contextual insights from captured or saved content material. As an illustration, while you snap a photograph, Moto AI works quietly within the background, pulling in important particulars and context concerning the picture. With only a easy voice command, you may revisit these recollections anytime.
Throughout its demonstration, Motorola showcased how Moto AI can function autonomously, mimicking human-like interplay with out the consumer needing to the touch the display screen—one thing different AI assistants, like Siri or Google Gemini, wrestle to attain. Customers can merely say instructions like “E book an Uber” or “Reserve dinner at my favourite restaurant,” and watch because the AI executes these duties proper earlier than their eyes, with an interface that easily mirrors typical human smartphone interactions.
Motorola’s revolutionary method integrates a number of AI fashions from OpenAI, Stability AI, and others, permitting the AI to excel in particular duties. The corporate goals to place Moto AI as a direct competitor to established leaders within the AI assistant area, together with Apple’s Siri and Google’s Gemini, each of that are closely investing in generative AI and client information safety. Nonetheless, a noteworthy facet of Moto AI is its steady recording of actions occurring throughout the telephone, enabling it to learn notifications, emails, and messages internally. So, is Moto AI really protected for customers?
Privateness vs. Comfort: How Safe Are These AI Methods?
Whereas the comfort of those AI assistants is obvious, the query of privateness stays a big concern. This might increase alarms that the telephone’s AI assistant is aware of extra concerning the consumer’s private life than themselves, even whether it is processed domestically. In response to client information privateness issues, Jeff Snow, Motorola’s Head of AI and Software program Experiences, emphasised that the AI assistant centralizes delicate consumer information to the telephone itself.
“We consider our AI-powered telephones will come throughout as a robust gadget for productiveness, because the AI doesn’t contact do not contact app information however puppeteer human actions by recording consumer preferences,” Snow instructed me. “Our method to AI is totally different, as a substitute of the consumer being the one inventive we’re enabling strategies to the customers from AI on how they will collaborate with the AI. We’re at the moment in Beta and plan to resolve extra consumer challenges by means of options earlier than launch.”
Lenovo and Motorola are at the moment centered on balancing AI-powered comfort with privateness safety. Their on-device AI techniques supply enhanced personalization, however public belief in AI assistants will hinge on customers’ confidence in information safety. Will customers embrace the comfort whereas not being utterly assured their private information is protected? The reply to this query might form the way forward for AI-driven client expertise.
