Davos 2026: Jensen Huang Calls for 'Trillions' in AI Infrastructure | January 26, 2026

Daily AI Blog

January 26, 2026

📋 Quick Takeaways

NVIDIA CEO Jensen Huang declares AI development requires “trillions of dollars” in infrastructure spending at Davos 2026
Trump administration commits to nuclear power expansion with 300 GW target by 2030 to fuel AI data centers
ChatGPT market share plummets from 86.6% to 64.6% while Google Gemini surges to 22% in one year
Waymo launches Miami robotaxi service, expanding to 6th US city with plans for 20+ cities in 2026
Congress advances AI Overwatch Act requiring approval for AI chip exports to China
DeepSeek prepares V4 model with enhanced coding capabilities ahead of Lunar New Year release
Physical AI emerges as dominant 2026 theme with 5-6x market potential versus agentic AI

🏛️ Davos 2026: AI Infrastructure Takes Center Stage

Jensen Huang: AI Requires “Trillions of Dollars” in Infrastructure Investment

Major Declaration: At the World Economic Forum in Davos, NVIDIA CEO Jensen Huang delivered a stark message to global leaders: artificial intelligence development will require “trillions of dollars” of spending on what he called “the largest infrastructure build-out in history.”

Key Statements:

Global data center power usage projected to surge from 55 GW to 84 GW within two years
AI is becoming “core infrastructure” for every economy
Current hyperscaler spending of hundreds of billions is just the beginning
Infrastructure costs will determine winners in the AI race

Strategic Context: Huang’s comments came during a conversation with BlackRock CEO Larry Fink, signaling that the world’s largest asset managers are aligning investment strategies around AI infrastructure as a multi-decade opportunity.

Source: Yahoo Finance | CNBC | World Economic Forum

Trump Administration Commits to Nuclear Power for AI Data Centers

Policy Shift: President Trump emphasized at Davos that the US is “going heavy into nuclear” to meet AI’s massive energy demands, touting executive orders signed in May 2025 targeting 300 gigawatts of new nuclear capacity by 2030.

Infrastructure Plans:

Department of Interior requiring tech companies to “pay their own way” for power infrastructure
Push for PJM utility to hold emergency power auctions with 15-year contracts for data center operators
Focus on accelerating regulatory approval for reactor construction
Increased domestic production of nuclear fuel

Trump’s Comments:

“We needed more than double the energy currently in the country just to take care of the AI plants… Instead of closing down energy plants, we’re opening them up.”

Industry Implications: The nuclear pivot addresses the critical energy bottleneck threatening AI expansion while positioning the US competitively against China in the AI infrastructure race.

Source: Yahoo Finance | World Economic Forum

Microsoft CEO Nadella: Energy Costs Will Determine AI Race Winners

Strategic Insight: Satya Nadella stated at Davos that energy and infrastructure costs will be the decisive factor in determining AI market leadership.

Key Quote:

“Are you a cheap producer of energy? Can you build data centers? What’s the cost of the silicon in the systems? We have got a new commodity—it is tokens. The job of every economy, and every firm in the economy, is to translate these tokens into economic growth.”

Market Reality: All major hyperscalers (Amazon, Google, Meta, Microsoft) face capacity constraints against strong AI demand and order backlogs. HSBC analysts note this constraint won’t change in 2026 due to long infrastructure build times.

Source: Yahoo Finance | Quartz

📊 AI Market Dynamics: ChatGPT vs Gemini

Seismic Shift: According to Similarweb data released January 22, ChatGPT’s worldwide traffic share declined to 64.6% as of January 16, 2026—down from 86.6% just twelve months earlier. This represents a 22-percentage-point loss in one year.

Market Share Changes (January 2025 → January 2026):

Platform	Jan 2025	Jan 2026	Change
ChatGPT	86.6%	64.6%	-22.0%
Gemini	5.3%	22.0%	+16.7%
Grok	—	3.5%	New
DeepSeek	—	3.3%	New
Claude	—	2.1%	—
Perplexity	—	1.9%	—

Key Developments:

Grok overtook DeepSeek for the first time (3.5% vs 3.3%)
Apple-Google partnership positions Gemini as foundation for Apple Intelligence
Google’s multi-year deal with Apple signals accelerating AI comeback

Strategic Implications: The rapid erosion of ChatGPT’s dominance demonstrates that first-mover advantage alone cannot sustain AI market leadership. Google’s integration strategy and ecosystem advantages are proving decisive.

Source: PPC Land | Axios

🚗 Autonomous Vehicles: 2026 “Year of the Robotaxi”

Waymo Launches Fully Autonomous Robotaxi Service in Miami

Major Expansion: Waymo officially opened its robotaxi service to the public in Miami on January 22, 2026, marking its 6th US city and extending its leadership position over competitors.

Service Details:

Coverage area: approximately 60 square miles of Miami’s urban core
Neighborhoods: Design District, Wynwood, Brickell, Coral Gables
Initial access: ~10,000 pre-registered users via waitlist
Future expansion: Miami International Airport later in 2026

2026 Expansion Plans:

Target: 20+ cities worldwide including Tokyo and London
US cities confirmed: Dallas, Houston, San Antonio, Orlando
Fleet expected to exceed 1,000 vehicles per major market

Competitive Context: Waymo provided 14 million fully autonomous paid rides in 2025 without any employees aboard, totaling 20 million over the past five years. Tesla has yet to offer a single fully autonomous robotaxi ride.

Source: Yahoo Finance - Autos | SF Standard | Electrive

Tesla Robotaxi Reality Check: Still No True Level 4 Autonomy

Current Status: Despite claims of “unsupervised” vehicles, Tesla’s Austin robotaxis still require remote safety monitors and are not classified as true SAE Level 4 autonomy.

Key Facts:

Tesla has reported 8 accidents to NHTSA even with safety employees monitoring
Service available in only 2 cities: Austin and San Francisco Bay Area
All rides require company employee presence (in-vehicle or remote)
Original goal of “half the US population” by end of 2025 missed significantly

Elon Musk’s Davos Claims:

“Tesla’s rolled out robotaxi service in a few cities, and will be very, very widespread in the US by the end of 2026.”

Industry Assessment: Carnegie Mellon professor Raj Rajkumar told media that Tesla is “way behind Waymo” in autonomous vehicle technology. The fundamental difference: Waymo uses redundant LiDAR, radar, and cameras, while Tesla relies solely on cameras and AI.

Source: CNN Business | Electrive | Futurism

2026 Predicted as Breakthrough Year for Autonomous Vehicles

Industry Consensus: Multiple analysts and executives predict 2026 will be the definitive “year of autonomous” driving with significant fleet expansion.

Fleet Projections:

Total US robotaxi fleet: 8,000 → 18,000+ vehicles (more than doubling)
Waymo: targeting 20+ cities globally
Tesla: aiming for 30 US markets (ambitious given current limitations)
Uber: robotaxi services in 10 markets by year end

New Market Entrants:

Zoox (Amazon): Paid robotaxi service launching H2 2026 in San Francisco
Uber: “Uber-exclusive robotaxi” partnership with Lucid Motors and Nuro
AVRide: Expanding through Uber app integration

Analyst View: Wedbush’s Dan Ives stated: “Tesla will successfully launch Robotaxis in over 30 cities in 2026 and start to scale volume production of Cybercabs starting the true autonomous era.”

Source: Yahoo Finance Video | Bloomberg

🛡️ AI Policy & Geopolitics

Congress Advances AI Overwatch Act to Restrict Chip Exports to China

Legislative Development: The House Foreign Affairs Committee advanced the AI Overwatch Act, a bill that would require congressional approval for AI chip export licenses to countries of concern within 30 days.

Key Provisions:

Revoke existing licenses for advanced AI chip transfers
Impose temporary ban until national security strategy submitted
Require both House Foreign Affairs and Senate Banking Committee approval
Exemptions for “trusted” US companies meeting security standards

Political Dynamics:

Bipartisan support from China hawks including Rep. Brian Mast (R-Fla.) and Rep. John Moolenaar (R-Mich.)
Directly challenges Trump administration’s plan to allow Nvidia H200 sales to China
Chinese regulators have separately instructed customs to block H200 imports

Chairman Mast’s Statement:

“Companies like Nvidia are requesting to sell millions of advanced AI chips, which are the cutting edge of warfare, to Chinese military companies like Alibaba and Tencent.”

Source: CNBC

🤖 Physical AI & Emerging Technologies

Physical AI Emerges as Dominant 2026 Theme

Industry Shift: Technology leaders at Davos and CES identified Physical AI—AI that takes physical form through robotics, autonomous vehicles, drones, and wearables—as the defining trend for 2026.

Market Projections:

EY estimates Physical AI market could be 5-6x the size of agentic AI within 5-6 years
Synopsys CEO: Physical AI arriving “much faster” than the expected 5+ year timeline
Jensen Huang: AI robotics is a “once-in-a-generation” opportunity for Europe

Key Developments:

Boston Dynamics & Google DeepMind announced major AI partnership at CES 2026
LG CLOiD smart home robot launched, powered by NVIDIA Jetson Thor
Meta Ray-Ban smart glasses gaining traction with neural interface features

Expert Assessment: “Robotics and physical AI are definitely going to pick up,” said IBM’s Peter Staar. “People are getting tired of scaling and are looking for new ideas.”

Source: CNBC | TechCrunch | IBM Think

DeepSeek Prepares New Flagship Model Ahead of Lunar New Year

Upcoming Release: Chinese AI lab DeepSeek is reportedly preparing to launch its V4 model with enhanced coding capabilities in mid-February 2026, coinciding with the first anniversary of its groundbreaking R1 reasoning model.

Recent Technical Advances:

mHC Architecture (Manifold-Constrained Hyper-Connections): New training method for improved model scaling
Engram Technique: Approach to build larger AI models with less powerful chips
Both innovations address US chip export restrictions

Strategic Significance:

DeepSeek’s R1 “Sputnik moment” in January 2025 shocked the industry by matching top competitors at fraction of cost
V4 expected to integrate new training efficiencies
Signals continued Chinese AI competitiveness despite hardware constraints

Analyst View: Counterpoint Research’s Wei Sun called DeepSeek’s recent papers a “striking breakthrough” demonstrating the company can “bypass compute bottlenecks and unlock leaps in intelligence.”

Source: Tech Wire Asia | South China Morning Post | SiliconANGLE

AI Agent Adoption Growing But Not Yet Autonomous

Enterprise Reality: While AI agents are being implemented across industries, experts at Davos emphasized that true autonomous operation remains elusive.

Current State:

Prosus: Operating 30,000 AI agents currently; CEO predicts agent-run companies within 5 years
Tencent: Companies moving from FOMO-driven pilots to production deployments
Most agents handling simple processes, not sophisticated workflows

Expert Caution:

“I think these agents are not autonomous. I don’t think we have reached the point where we can replace a human employee.” — Uljan Sharka, CEO of Domyn

Slack/Salesforce Prediction: Ryan Gavin forecasts “2026 will be the year of the lonely agent”—companies will deploy “hundreds of agents per employee,” but most will sit idle like unused software licenses.

Source: CNBC | Axios

Elon Musk Predicts AI Surpasses Human Intelligence in 2026

Bold Prediction: At Davos 2026, Elon Musk predicted that AI could surpass human intelligence “as soon as this year” and that “AI will be smarter than all of humanity collectively” by 2030.

Additional Musk Claims:

Tesla has “solved” autonomous driving
Plans for solar-powered AI data centers in space
Tesla robotaxis will be “very, very widespread” in US by end of 2026

Context: Musk’s predictions come amid a pattern of ambitious timelines that have frequently been missed. His 2019 claim of one million robotaxis by 2020 did not materialize, and Tesla continues to lag Waymo in actual autonomous deployments.

Source: CNBC | World Economic Forum

📈 Market Impact Analysis

The convergence of infrastructure demands, market shifts, and policy tensions at Davos 2026 signals a critical inflection point for the AI industry:

Infrastructure Economics: Jensen Huang’s “trillions of dollars” declaration and Trump’s nuclear pivot underscore that energy and infrastructure costs have become the primary constraint on AI development—not algorithm advancement.

Competitive Realignment: ChatGPT’s 22-percentage-point market share loss to Gemini demonstrates that distribution and ecosystem integration (Google’s Apple partnership) now matter more than first-mover advantage.

Autonomous Vehicle Inflection: With Waymo’s 14 million autonomous rides versus Tesla’s zero, 2026 will determine whether Tesla’s camera-only approach can catch up to LiDAR-based systems—or if Musk’s predictions will again prove premature.

Geopolitical Fracture: The AI Overwatch Act’s advancement signals growing congressional resistance to AI chip exports despite administration support, potentially fragmenting the global AI supply chain.

Physical AI Emergence: The industry’s pivot toward robotics, autonomous vehicles, and wearables suggests the next AI value creation wave will occur in the physical world, not just digital interfaces.

🔮 Looking Ahead

Key Trends to Watch:

Hyperscaler Q4 earnings (late January) for AI capex guidance
DeepSeek V4 model release (mid-February)
AI Overwatch Act congressional progress
Waymo expansion to Dallas, Houston, San Antonio, Orlando
Apple Intelligence with Gemini integration launch timeline

Stay Updated: Follow us for comprehensive daily AI news, policy analysis, and market tracking.

Last Updated: January 26, 2026, 11:59 PM CST