|
|
|
|
|
|
|
|
|
|
The Morning Download: Is Google the Next Google?
|
|
By Steven Rosenbush | WSJ Leadership Institute
|
|
|
|
|
|
|
What's up: Chip leaders at CES talk jobs, taxes; LinkedIn is so hot right now; the fight over making data centers power down to avoid blackouts.
|
|
|
|
|
|
Attendees at a Google conference last year tried out activities highlighting Gemini AI. Camille Cohen/AFP/Getty Images
|
|
|
|
|
|
Good morning. Google kicked off a historic AI boom in 2017 when eight of its researchers published “Attention Is All You Need,” a paper that laid out a new approach to deep learning. The company lost the initiative, however, to OpenAI, formed in part out of fear that Google would end up dominating all of AI.
The WSJ’s Katherine Blunt explores how Google, under the leadership of CEO Sundar Pichai, regained its momentum. With the launch of AI Overview, Google defied predictions that ChatGPT would bring about the end of its core market in search. It combined its DeepMind and Google Brain labs under the leadership of Demis Hassabis, and with last fall’s launch of the Gemini 3 models regained the high-ground of development. Now it was OpenAI’s turn to go Code Red.
Google’s comeback may be just getting started. Google is a reminder of just how hard it is to lead in AI development, let alone maintain that edge. Those who win, win big. As the WSJ noted last month, Apple is reportedly testing a version of Gemini to power a new version of Siri. And Reuters reported Monday that Samsung Electronics plans to double the number of mobile devices with "Galaxy AI" features largely powered by Gemini.
Product, the final frontier. When it comes to science-driven innovation, Google still functions like a startup. But the reality is that it’s an enormous enterprise and friction and bureaucracy are an institutional risk. Turning breakthrough ideas into breakthrough products presents its own challenges. However, the success of AI Overview shows that Google has the capacity for success on this level as well.
Take-away for enterprise leaders. If there’s a teachable moment in Google’s return to glory, it is the company’s ability to recognize a crisis and respond accordingly. When ChatGPT hit, Google didn’t panic. No, it freaked out, and that was the right thing to do. OpenAI responded in kind when it declared Code Red. What if automakers, entertainment companies and other enterprises had dropped their defenses and leaned into crisis when they faced existential threats? In a world of technology-driven change, companies need to sharpen their animal instincts and know when and how to respond to danger.
|
|
|
|
|
Content from our sponsor: Deloitte
|
|
|
AI Agent Orchestration: Unlocking Exponential Value
|
|
Autonomous enterprise AI agents can be transformational, but realizing their full value increasingly depends on thoughtful orchestration. Read More
|
|
|
|
|
|
|
|
|
|
|
Chip Leaders at CES Talk Jobs, Taxes...and Chips (Of Course)
|
|
|
|
|
|
Jensen Huang, chief executive of Nvidia. Bridget Bennett/Bloomberg News
|
|
|
|
|
|
Advanced Micro Devices Chief Executive Lisa Su tells CNBC that AI is a big factor in hiring at the chip maker, but not as a reason to hire less: Quite the opposite. “Frankly, we’re growing very significantly as a company, so we actually are hiring lots of people, but we’re hiring different people,” she said. “We’re hiring people who are AI forward.”
Nvidia CEO Jensen Huang in a discussion with Bloomberg Television said he's not worried about a potential tax on billionaires in California, breaking from some ultrawealthy residents who have spoken out against the proposal. The ballot initiative would impose a one-time, 5% tax on the assets of those with net worths above $1 billion.
“I’ve got to tell you, I have not even thought about it once,” Huang said Tuesday. “We chose to live in Silicon Valley, and whatever taxes I guess they would like to apply, so be it. I’m perfectly fine with it.”
|
|
|
And Huang, in a separate question and answer at the Consumer Electronics Show, said Chinese demand for Nvidia’s H200 advanced AI processors is “quite high.”
“We’ve fired up our supply chain, and H200s are flowing through the line,” Huang said Tuesday.
The Trump administration made the controversial decision to approve the chips’ sale in China.
|
|
|
|
|
|
|
Artificial intelligence is no longer just a software story. It’s reshaping the real economy, from surging demand for memory chips, to a funding squeeze that’s changing how companies raise money, to growing strain on the power grid.
|
|
The Fight Over Making Data Centers Power Down to Avoid Blackouts
|
|
|
|
|
|
A surge in data-center development has pushed electricity prices higher and increased the threat of blackouts in some parts of the U.S. Ana Elisa Sotelo for WSJ
|
|
|
|
|
|
With data centers’ rising power demands straining parts of the U.S. grid, operators are asking tech companies to make accommodations until new transmission lines and power plants are built to meet demand.
|
|
|
Some grid operators, like PJM Interconnection, have proposed either requiring or encouraging data centers to stop using it when there is a risk of blackouts, either by powering down or switching to backup electricity supplies.
New rules. The WSJ reports that Southwest Power Pool, which operates part of the grid stretching from the Texas Panhandle to North Dakota, plans to begin offering data-center operators the option of “conditional” service, allowing them to connect to the grid faster with the understanding that they might get cut off when supplies tighten.
Data centers are pushing back. A trade group has called such proposals discriminatory, arguing that data centers need to run constantly to train and run AI models, but also to serve cloud-computing needs.
The Google way. The tech giant for years has been working to determine how it can use less power during times of strain on the grid. The company has pilot programs with several utilities in which it reduces power use within its data centers in exchange for payment, a tactic known as demand response.
|
|
|
|
Venture-Capital Fundraising Falls
|
|
|
|
|
Fundraising for U.S. venture-capital firms dropped 35% in 2025, the most anemic stretch in at least six years, with money flowing primarily to the most trusted investment firms as companies stay private longer.
|
|
|
The slowdown reflects a continuing liquidity crunch and helps explain why cash-intensive AI businesses are looking beyond traditional venture firms for capital, such as the deep-pocketed sovereign-wealth funds, family offices and hedge funds, the Journal reports.
|
|
|
More fund-raising news.
xAI, Elon Musk’s artificial intelligence company, completed a $20 billion Series E funding round, surpassing its $15 billion target, with investors including Qatar Investment Authority, Valor Equity Partners, Nvidia and Cisco Investments, WSJ reports. The company is currently facing probes in Europe and elsewhere for its Grok chatbot that recently has been generating explicit deepfakes, some involving children, CNBC reports.
|
|
|
|
|
|
|
|
|
Investors are paying close attention to flash memory and memory chips, essential components of AI system performance. On Tuesday, shares of flash-memory maker Sandisk soared more than 27% to a record $349.63. While it wasn’t clear what sparked the rise, Morningstar analyst William Kerwin linked the move to Nvidia’s CES keynote on Monday, Barron’s reports, where the company discussed a new memory storage platform for its Rubin chip.
|
|
|
|
LinkedIn, Favorite Social Media of the Professional Class, Is Suddenly Hot
|
|
|
Ignore the corporate buzzwords and stories about overcoming workplace adversity. The social-media site has emerged as one of the last places of authentic online identity.
|
|
|
|
|
The secret is a founding principle that curbs toxicity, the WSJ's Stu Woo reports. Users are required to provide their real names. While not immune to misinformation and scams, LinkedIn in recent years has lured people leaving X and Facebook as content moderation and fact-checking there declined.
Notes the WSJ: "Many concluded it was worth trading rage bait on other platforms for earnest monologues about why getting laid off was a blessing in disguise."
Revenue jumped to $17 billion in 2025 from $7 billion in 2020, and membership doubled to 1.3 billion. Users stick around too: Americans checking LinkedIn more than once a day climbed to 4.7% last year from 3.9% in 2020, according to research firm GWI.
|
|
|
|
|
|
|
|
|
Meta’s former chief AI scientist Yann LeCun criticized the company’s AI leadership, calling its chief AI officer, Alexandr Wang, young and inexperienced, and warned of a potential staff exodus, FT reports. LeCun late last year announced he was leaving the company to start his own venture, Advanced Machine Intelligence Labs, focused on world models. Wang joined Meta in 2025 after it bought a 49% stake in Scale AI.
|
|
|
|
Everything Else You Need to Know
|
|
|
Secretary of State Marco Rubio told lawmakers that recent administration threats against Greenland didn’t signal an imminent invasion and that the goal is to buy the island from Denmark, according to people familiar with the discussions. (WSJ)
President Trump intends to meet representatives of the three largest U.S. oil companies and other domestic producers at the White House on Friday to discuss making significant investments in Venezuela’s oil sector, according to people familiar with the matter. (WSJ)
U.S. forces moved to seize two tankers on Wednesday, including one that had fled the U.S. blockade of sanctioned vessels near Venezuela that was being escorted by a Russian submarine in the eastern Atlantic, according to U.S. officials. (WSJ)
The U.K. and France will set up military hubs across Ukraine and build protected facilities to produce weapons and military equipment for the country if a cease-fire agreement is reached between Kyiv and Russia, British Prime Minister Keir Starmer said Tuesday. (WSJ)
|
|
|
|
|
Content From Our Sponsor: DELOITTE
|
|
13 Predictions for 2026: The AI Gap Narrows—but Likely Persists
|
|
Deloitte Global predicts that the gap between the promise and reality of AI will continue to narrow in 2026 as further developments help propel the technology toward broader adoption. Read more.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|