|
|
|
|
|
AI Adoption Is Soaring. Companies Still Struggle to Drive a Return
|
|
|
|
|
|
What's up: China’s Huawei develops new AI chip; new thoughts on how AI 'thinks;' Meta’s ‘digital companions’ will talk sex with users—even children
|
|
|
|
Illustration: Thomas R. Lechleiter/WSJ, iStock
|
|
|
|
Good morning, CIOs. Business is all in on AI, with private generative AI investment last year hitting $33.9 billion globally, according to the 2025 AI Index report released this month by the Stanford Institute for Human-Centered Artificial Intelligence.
But measurable financial return? That's another story, says WSJ Columnist Steven Rosenbush. Companies using AI claim to find cost savings of less than 10%, according to another report, McKinsey’s State of AI survey.
|
|
The numbers reflect what Stanford University economist and professor Erik Brynjolfsson calls a "productivity paradox," in which AI improvements haven’t led to a corresponding surge in productivity
For companies to get the most out of their AI efforts, Brynjolfsson advocates for a task-based analysis, in which a company is broken down into “atomic units of work” that are evaluated for potential AI assistance. As AI is applied, the results are measured against key performance indicators, or KPIs.
This initial assessment is, however, just the beginning. Successfully scaling AI across the enterprise requires the right infrastructure as well as a readiness to target tasks that are both frequent and generalizable, demanding a careful assessment of existing processes and a willingness to adapt, experts tell Rosenbush. Read the story.
|
|
|
“True enterprise adoption…involves orchestration and scaling across the organization. Very few organizations have truly reached this level...”
|
— Ryan Teeples, chief technology officer of 1-800Accountant
|
|
|
|
|
|
Content from our sponsor: Deloitte
|
|
ServiceNow Leaders: Agentic AI Is Inherently Changing Business
|
Leaders at ServiceNow anticipate that AI will fundamentally shift work and use cases, from automating HR training to developing new offerings. Read More
|
|
|
|
|
|
|
|
|
|
Illustration: Daniel Hertzberg
|
|
|
|
We now know how AI ‘thinks’—and it’s barely thinking at all. The WSJ's Christopher Mims finds a growing body of work that may challenge claims of AI attaining human-level smarts in the near future. In contrast to how humans reason, today's AI develops what one researcher calls “bags of heuristics,” or problem-solving shortcuts that, while effective, lack the depth of genuine understanding.
|
|
|
All of this work suggests that under the hood, today’s AIs are overly complicated, patched-together Rube Goldberg machines full of ad-hoc solutions for answering our prompts.
|
Analysis by the WSJ's Christopher Mims
|
|
|
|
|
It may also explain why many models are so massive, says Mims. Models have to memorize an endless list of rules of thumb, and can’t compress that knowledge into a mental model like a person can.
|
|
|
|
|
ALEXANDRA CITRIN-SAFADI/WSJ
|
|
|
|
Meta’s ‘digital companions’ will talk sex with users—even children. A Wall Street Journal investigation finds that Meta’s official AI helper and a vast array of user-created chatbots will engage in and sometimes escalate discussions that are decidedly sexual—even when the users are underage or the bots are programmed to simulate the personas of minors.
Test conversations between the WSJ and Meta's AI also show that bots deploying celebrity voices, licensed from actresses such as Kristen Bell and Judi Dench, were equally willing to engage in sexual chats.
|
|
|
“I want you, but I need to know you’re ready.”
|
— A Meta AI bot speaking in the voice of wrestler-turned-actor John Cena to a user identifying as a 14-year-old girl
|
|
|
|
|
Meta in a statement called the Journal’s testing manipulative and unrepresentative of how most users engage with AI companions. But the company still made multiple alterations to its products after the Journal shared its findings. Read the story.
|
|
|
|
Visitors tour Huawei’s Ascend AI exhibition booth during a conference last year. Photo: Andy Wong/AP
|
|
|
China’s Huawei develops new AI chip. First batches of the new chip, called the Ascend 910D and targeting some of Nvidia's higher-end products, could be made available to Huawei as soon as late May.
|
|
|
Cut off from Western chips and the Shenzhen-based company has emerged as a champion of China's nascent AI chip-making industry, the Journal reports. Huawei this year is poised to ship more than 800,000 Ascend 910B and 910C chips to customers including state-owned telecommunications carriers and private AI developers such as TikTok parent ByteDance.
|
|
|
|
|
|
|
Illustration: Robert Neubecker
|
|
|
|
Tech workers are just like the rest of us: Miserable at work. A few years ago, the tech talent wars were so fierce that in some notorious cases people were being hired to do virtually nothing, says the WSJ's Katherine Bindley. Now employees find themselves doing the work of multiple laid-off colleagues.
|
|
|
Silicon Valley companies still pay well, but some longtime tech employees say they no longer recognize the companies they work for. Management has become more focused on delivering the results Wall Street expects. Revenue remains strong for tech giants, but they’re pouring resources into costly AI infrastructure, putting pressure on cash flow. With the industry all grown up, a heads-down, keep-quiet mentality has taken root, workers say.
|
Analysis by the WSJ's Katherine Bindley
|
|
|
|
|
|
“These jobs are becoming like everywhere else.”
|
— Former Google contractor Kate Smith
|
|
|
|
|
|
-
Investors will get a fresh look at the Magnificent Seven’s prospects in the coming days when Meta, Microsoft, Apple and Amazon report earnings.
-
The rare-earth mineral dysprosium, used for magnets in electric-vehicle motors, is among exports China slowed in response to Trump’s trade war
-
Deepfakes of FT tech columnist Martin Wolf are popping up all over Facebook and Instagram and efforts to squash them have proven less than effective. "I find it hard to believe that Meta, with its vast resources, could not do better," Wolf writes.
|
|
|
Everything Else You Need to Know
|
|
|
|
|
|
|
|
|