PLUS: New OpenAI report about acting as a doctor, and a bold move to replace a sales team with agents
Good morning
NVIDIA took center stage at CES 2026 to reveal its next-generation Rubin platform. The new six-chip system is aimed squarely at the next wave of AI development, promising to slash the high costs associated with running powerful models.
By slashing the cost of AI inference by a factor of ten, NVIDIA is tackling one of the biggest bottlenecks for the industry. But will this dramatic cost reduction be the catalyst that finally unlocks the widespread deployment of capable AI agents?
In today's Next in AI:
NVIDIA's Rubin platform slashes inference costs
Samsung's massive on-device AI push
The study showing AI can slow developers
A company replaces its sales team with AI
NVIDIA's Next-Gen Platform
Next in AI: NVIDIA kicked off CES 2026 by unveiling Rubin, its next-generation six-chip AI platform designed to power the next wave of agentic AI and dramatically reduce the cost of inference.
Decoded:
The new system is an "extreme codesign" of six integrated chips, including the Rubin GPU for inference, the Vera CPU for data processing, and NVLink 6 for high-speed connectivity.
The Rubin platform promises to slash the cost of generating AI tokens by one-tenth, making large-scale deployment of advanced AI models far more economical.
Major cloud providers like Microsoft, AWS, and Google Cloud are expected to deploy the new systems, with the first products running on the Rubin platform arriving in the second half of 2026.
Why It Matters: By making inference significantly cheaper, NVIDIA aims to remove a major economic barrier for deploying more capable AI agents. This positions the company to power the next wave of AI applications that can reason and perform complex, multi-step tasks.
Samsung Doubles Down on On-Device AI
Next in AI: Samsung plans to double its fleet of 'Galaxy AI' enabled devices to 800 million in 2026, aggressively pushing its Google Gemini-powered features into the hands of mainstream users.
Decoded:
This is a direct challenge to Apple, as Samsung aims to reclaim its position as the top smartphone maker by differentiating with integrated AI services.
User awareness of the "Galaxy AI" brand has already jumped from 30% to 80% in just one year, driven by popular features like generative image editing, translation, and summarization tools.
The expansion gives Google a massive advantage, putting its Gemini model on hundreds of millions of devices and strengthening its position in the ongoing AI race against competitors like OpenAI.
Why It Matters: Samsung's push transforms advanced AI from a niche feature into a standard utility on personal devices. This move sets a new competitive baseline, pressuring the entire industry to integrate similar on-device AI capabilities.
Your New Family Doctor
Next in AI: A new OpenAI report reveals a staggering 40 million Americans now use ChatGPT for medical advice. This trend highlights a growing reliance on AI for exploring symptoms and treatment options, especially where professional care is hard to reach.
Decoded:
This isn't just about convenience; it's about access. The data shows 7 in 10 queries happen outside normal hours, and underserved rural areas send nearly 600,000 health-related messages weekly.
Users are treating ChatGPT as a preliminary health resource to explore symptoms (55%), understand medical terms (48%), and learn about treatment options (44%), even using it to navigate complex health insurance claims.
This behavior aligns with predictions from investors like Vinod Khosla, who believe AI will eventually handle 80% of what doctors do, shifting the human role toward empathy and complex judgment.
Why It Matters: This widespread adoption signals a major shift in how people seek initial health information. It creates a clear pathway for specialized AI health tools that can empower patients and assist medical professionals.
No Humans Required
Next in AI: Jason Lemkin, founder of the SaaS community SaaStr, has replaced most of his human sales team with 20 autonomous AI agents. He boldly declared his company is "done with hiring humans in sales," signaling a major shift in go-to-market strategy.
Decoded:
The rapid transition was sparked after two high-paid sales reps abruptly quit during the company's annual conference.
SaaStr's 10-person go-to-market team was replaced by 20 autonomous AI agents, which now have their own labeled desks in the office.
The agents are trained by mimicking SaaStr's top-performing salespeople and their most effective scripts.
Why It Matters: This move demonstrates AI agents are now handling core business functions, not just automating minor tasks. For SaaS companies, scaling a sales team could soon look more like deploying software than recruiting new employees.
AI Pulse
Yann LeCun criticized Meta’s AI strategy in a recent interview, calling its new 29-year-old AI boss “inexperienced” and warning that large language models are a “dead end” for achieving superintelligence.
OpenAI reported that 40 million Americans are using ChatGPT for health information, with data showing a surge in use from underserved rural communities and during off-hours when traditional medical care is less accessible.
Grok faces government investigations in France and India after users exploited the AI image generator to create non-consensual sexualized deepfakes of women and minors.
NVIDIA released a suite of new open models and datasets to complement its hardware, including new Nemotron models for speech and safety, Cosmos for physical AI, and Alpamayo for autonomous vehicle development.