To The Moon

By Krystal Hu

Jan 7 (Reuters) - (Artificial Intelligencer is published every Wednesday. Think your friend or colleague should know about us? Forward this newsletter to them. They can also subscribe here or email me to share any thoughts.)

Happy New Year and welcome back! I hope you had a joyful, restful holiday.

For our global tech editor Ken Li, the AI boom hit home while shopping for a Christmas gift: his 13-year-old son’s first gaming PC. What should have been a straightforward purchase turned into a pricing nightmare. Those gaming rigs sporting the latest hardware were 60% more expensive than they were merely a few weeks ago, revealing the personal cost of the memory chip shortage brought on by the AI boom and now squeezing the consumer supply chain.

The sentiment was echoed inside the buzzing hotels in Las Vegas this week, from the annual Consumer Electronics Show, where Nvidia NVDA.O CEO Jensen Huang talked about the exploding need for data storage, lifting the stock price of long-overlooked names from Sandisk SNDK.O to Micron MU.O as investors hunted for the next choke point in the AI supply chain.

My colleagues have been tracking this global shortage and talking to industry insiders across five countries since last year. We’ll decode what this means for consumers and the AI buildout. Plus, we’ve got the latest data on global VC funding and the ever-growing pie of AI. Scroll on.

OUR LATEST REPORTING IN TECH & AI:

• Musk's xAI raises $20 billion in upsized Series E funding round

• Nvidia CEO Huang says next generation of chips is in full production

• Microsoft taps major US electric grid operator to modernize the Midwest power system

• European Commission calls Grok's sexualised AI photos 'illegal,' Britain demands answers

• Hyundai Motor Group plans to deploy humanoid robots at US factory from 2028

• SoftBank completes $41 billion investment in OpenAI, deepening bet on AI

A NEW KIND OF CHIP SHORTAGE (NOT NVIDIA)

The AI boom has an unexpected side effect, and it's heading for your wallet. The price of your next smartphone or PC is likely going up—not because of flashy new AI features, but because of a global scramble for one of the most basic components inside: memory chips.

What we’ve been hearing from more than 40 industry insiders in South Korea, Japan and the U.S. is consistent: tech companies are scrambling for memory. Microsoft MSFT.O, Google and ByteDance are racing to lock in supply from the three largest memory chip makers SK Hynix, Samsung and Micron.

The squeeze spans nearly every type of memory — from humble flash chips used in USB drives and smartphones to high-bandwidth memory $(HBM)$. HBM acts like a multi-lane highway that feeds data to power-hungry AI chips, and prices for some types have more than doubled since February.

Chipmakers are naturally following the money and in doing so, they’re starving everything else. Producing HBM for AI servers is more profitable than making conventional memory for phones, laptops and TVs. As a result, manufacturers are diverting capacity toward AI at a time when they still can’t produce enough high-end chips to keep up with demand.

Consider the scale. OpenAI, for example, signed supply agreements with Samsung 005930.KS and SK Hynix 000660.KS last October for its Stargate project, which could require up to 900,000 wafers per month by 2029 — roughly double today’s global HBM output.

Consumers are already feeling the crunch. In Tokyo’s electronics district of Akihabara, retailers have begun restricting memory purchases to curb hoarding. Samsung has warned that the shortage affects everything from phones to TVs and home appliances. Chinese smartphone makers Xiaomi 1810.HK and Realme said they may have to raise prices or focus on selling more premium models. IDC and Counterpoint now expect the global smartphone market to shrink next year as higher memory costs filter through.

Memory makers are expanding, but relief won’t come quickly. SK Hynix has said its 2026 production is sold out, while Samsung says customers have already lined up for next year’s HBM. New factories for conventional memory won’t meaningfully come online until 2027 or 2028.

Memory has always been a cyclical business, swinging between brutal busts and euphoric booms. What’s different this time is that AI is pulling the cycle forward, draining supply from everyday devices to feed data centers. There’s also a longer-term risk baked into this scramble. The industry hasn’t seen meaningful unit growth in years, yet it’s now being asked to absorb unprecedented capital spending to serve AI workloads. New fabs cost tens of billions of dollars and take years to come online, which means today’s shortage could easily become tomorrow’s oversupply if AI demand slows or shifts.

For now, the only certainty is that the AI gold rush is on, and the price of admission might just be hidden in the cost of your next gadget.

CHART OF THE WEEK:

AI isn't just a hot sector; it's becoming the only sector that matters for many venture capitalists. Almost two-thirds of global venture capital funding in 2025 went to AI companies, continuing a trend that began with the breakout of ChatGPT, according to the latest data from PitchBook. This AI funding boom helped propel global VC investment to its second-highest annual deal value on record, just $28.8 billion shy of the 2021 peak.

To put it in numbers, annual AI investment has surged from just over $70 billion to more than $210 billion in recent years, with deal counts up 24% over that period. Unlike the broad-based funding surge in 2021, this wave has all the capital concentrated into hyper-focused AI bets — large model developers, infrastructure providers, and AI-native platforms — while funding for non-AI startups has slowed to a trickle.

(Reporting by Krystal Hu; Editing by Lisa Shumaker)

((krystal.hu@thomsonreuters.com, +1 917-691-1815))