AI News Archive: May 1, 2026 — Part 12
Sourced from 500+ daily AI sources, scored by relevance.
- ChatGPT's goblin obsession may be hilarious, but it points to a deeper problem in AI training
A faulty reward signal during training caused ChatGPT models to start dropping goblins, gremlins, and other mythical creatures into their answers at a surprising rate. OpenAI says it's an example of how small, poorly tuned training incentives can produce unexpected side effects. The article ChatGPT's goblin obsession may be hilarious, but it points to a deeper problem in AI training appeared first on The Decoder .
- There’s a Fascinating Reason Why OpenAI’s Latest Model Is Obsessed With Goblins and Gremlins
The OpenAI model’s focus on creatures like goblins and trolls has a very nerdy explanation. Here’s what happened.
- OpenAI explains why ChatGPT suddenly loved goblins
OpenAI published a blog post explaining why a recent ChatGPT model was so obsessed with working goblins into answers.
- Central Texas is adding 70+ data centers. What it means for your health
Central Texas is adding 70+ data centers. What it means for your health Houston Chronicle
- 69% Of Workers Are Disengaged In The AI Era. A CEO Explains Why
A Gallup survey reveals that more than two-thirds of workers are disengaged. Here's why, and how to change that while adopting AI in the workforce.
- Mark Zuckerberg says most AI agents don't pass the 'mother' test
Mark Zuckerberg says most AI agents don't pass the 'mother' test Business Insider
- First Take: Appian Turns to Modernization as AI-Native Tools Threaten the Low-Code Market
First Take: Appian Turns to Modernization as AI-Native Tools Threaten the Low-Code Market Gartner
- Inside VentureOne: How Abu Dhabi’s Venture Builder Is Advancing Purpose-Driven Deep Tech
Inside VentureOne: How Abu Dhabi’s Venture Builder Is Advancing Purpose-Driven Deep Tech Entrepreneur Middle East
- Why Big Tech's $700 billion AI splurge is misleading
Why Big Tech's $700 billion AI splurge is misleading Business Insider
- Big Tech Q1 2026 earnings: Key takeaways on AI spending, cloud growth, and more
Big Tech Q1 2026 earnings: Key takeaways on AI spending, cloud growth, and more
- Big tech's AI spending balloons to $725 billion this year
Big tech keeps pouring more money into AI data centers, chips, and infrastructure. According to the Financial Times, Google, Amazon, Microsoft, and Meta have a combined budget of around $725 billion for next year. The article Big tech's AI spending balloons to $725 billion this year appeared first on The Decoder .
- US giants dominate AI spending over Chinese rivals
China should not be judged against US cloud giants alone, as local firms face curbs on advanced US chips.
- Google Is the Only Big Tech Company Making Its AI Spending Work
Google Is the Only Big Tech Company Making Its AI Spending Work Barron's
- Big Tech capex ranked: What Alphabet, Amazon, Apple, Meta, and Microsoft are spending as AI investment surges
For years, it was common for even the biggest tech companies to have annual capital expenditures, or capex, in the single- to low-double-digit-billion range. You might have heard a tech company say it planned to spend $9 billion, $15 billion, or even $25 billion on research, development, and other costs in the upcoming fiscal year. But lately, capital expenditures at the largest tech companies have been off the charts, with some companies now regularly forecasting single-year capex in the hundreds of billions. The driving factor for this is, of course, artificial intelligence (AI). Some of the biggest names in tech are throwing previously unthinkable sums behind AI development in an attempt to become the king of artificial intelligence down the road. This week, investors received an update on capex from five major tech companies—Alphabet, Amazon, Apple, Meta, and Microsoft—all of which reported their latest earnings. Here’s what they said they expect to spend on capex during their curr
- Army plans fast follow-up to AI cyber wargame with industry: Officials
Army plans fast follow-up to AI cyber wargame with industry: Officials Breaking Defense
- The White House rethinks its Anthropic fight
PLUS: Stress test business ideas with Perplexity
- Elon Musk says xAI partly used OpenAI models in AI training
In a high-profile hearing in a California federal court, Elon Musk said it was “partly” true that xAI had used OpenAI’s models in the development of its own Artificial Intelligence (AI) systems. The exchange came during the ongoing legal battle between Musk and OpenAI. Under questioning by OpenAI lawyer William Savitt, Musk was asked whether xAI had used distillation techniques involving OpenAI models. Musk first said such practices were common across the AI industry. When pressed further, he reportedly replied, “Partly.” According to Reuters, Musk testified that xAI used OpenAI to train its models. Technology publications including Wired and The Verge reported the specific distillation exchange, where Musk said it was “partly” true that xAI had used OpenAI models in its AI training. xAI is best known for Grok, its AI chatbot. The testimony has brought renewed attention to model provenance, intellectual property (IP) exposure, and transparency across the AI supply chain. For enterprise
- Musk Admits xAI Distilled OpenAI Models to Train Grok Under Oath
Elon Musk admits to using OpenAI models to train Grok
- Google Deepmind's "AI co-clinician" beats GPT-5.4 in blind doctor tests but still trails experienced physicians
Google Deepmind is building an "AI co-clinician" to help doctors care for patients. The system shows promising results in simulation studies but still trails experienced physicians. The research also shows why ChatGPT's voice mode isn't ready for serious tasks, let alone medical consultations. The article Google Deepmind's "AI co-clinician" beats GPT-5.4 in blind doctor tests but still trails experienced physicians appeared first on The Decoder .
- Philips doubles down on AI to offer better care for more people
SINGAPORE — Philips on Wednesday unveiled a new artificial intelligence-backed patient monitoring innovation strategy, underlining its commitment to help hospitals improve health care coordination to offer better care for more people and more effectively cope with projected workforce shortages worldwide. “We need to start with where the pain sits in health care,” Stephanie Sievers, managing director of Philips Asia Pacific, told reporters on the sidelines of the company’s Innovation Summit in Si
- Paging Dr. AI to the ER? Artificial intelligence shows promise in emergency room diagnosis
Paging Dr. AI to the ER? Artificial intelligence shows promise in emergency room diagnosis CBC
- Paging Dr. AI to the ER? Artificial intelligence shows promise in emergency room diagnosis
Paging Dr. AI to the ER? Artificial intelligence shows promise in emergency room diagnosis CBC
- AI beats doctors in diagnoses, but human judgment still key: study
AI beats doctors in diagnoses, but human judgment still key: study CBC
- Harvard Researchers Say Your Next ER Diagnosis May Come From AI—and It Could Be More Accurate Than Human Doctors
A new peer-reviewed study found AI diagnosed emergency patients more accurately than human doctors. Researchers believe it could reshape the future of medicine.
- What Are the Chances AI Will Give You Accurate Health Advice? 50/50, Says a New Study
If you wouldn’t trust a doctor to decide your diagnosis by flipping a coin, you probably shouldn’t trust AI for health advice either.
- How artificial intelligence is meeting the patient first
How artificial intelligence is meeting the patient first Healthcare IT News
- AI nailed emergency diagnoses better than doctors in Harvard trials
A Harvard trial found OpenAI’s o1 model beat doctors at emergency triage diagnoses, showcasing AI’s promise as a second-opinion tool in hospitals.
- AI Can Spot Pancreatic Cancer Years Before Diagnosis, Study Finds
Catching this deadly disease while it's treatable can save lives. ScienceAlert stories are written, fact-checked, and edited by humans, never generated by AI. Don't miss a story, subscribe here.
- On the Ground in China’s Humanoid Robotics Moment
China’s Humanoid Robots Are Moving Fast. Paul Triolo Explains Why.
- 1X shows off Neo humanoid robot helping humans make more of its kind
1X is moving toward scaled humanoid robot production, and its latest demo shows Neo already playing a role on the factory floor.
- AI race squeezes global memory supply
AI race squeezes global memory supply marketplace.org
- Mark Zuckerberg says Meta's layoffs are driven by surging AI spending. And he won't rule out more cuts
The roughly 8,000 positions being eliminated represent about 10% of Meta's global workforce
- Zuckerberg blames Meta layoffs on AI costs, says “compute and infrastructure” and “people oriented things” are biggest financial drain right now
Meta's capex rises as the company spends big on AI data centers, and it's having to reduce headcount to become more agile.
- Zuckerberg says Meta layoffs tied to AI spending, won't rule out future cuts
Mark Zuckerberg says Meta's plan to cut about 10% of its workforce is tied to increased AI investment, while leaving the door open to additional layoffs.
- UK job seekers lament menace of unnatural AI Interviews
UK Job Seekers Say Artificial Intelligence is Ruining the Interview Process
- AI is turning every story into raw material
There’s an idea in AI called “liquid content.” It typically refers to the idea of morphing the facts, ideas, and expressions from one medium to another. The most well-known example is a feature within Google’s NotebookLM : Once you’ve filled a folder with various kinds of data, it can whip up a podcast about that data, enlisting a couple of cheery AI-generated voices to give you an overview, analysis, or debate. Taken to its logical extreme, liquid content suggests a future for media companies where what you create is repurposed across any and all formats. Making a podcast? With the right tools and prompting, in mere minutes, it can be reimagined as a series of clips, a feature article, or even an interactive presentation. And if you’re a traditional news publisher, all that content can serve as raw material for videos, which you may have dismissed in previous eras as too expensive to produce. This isn’t theoretical anymore. I recently attended a couple of industry conferences—the NAB
- Toilet Maker Spikes in Value as It Flushes Money Into AI
The company just dropped a bomb. The post Toilet Maker Spikes in Value as It Flushes Money Into AI appeared first on Futurism .
- AI actors and writers not eligible for Oscars - Academy
Actors created with artificial intelligence will not be eligible for an Oscar, the Academy of Motion Picture Arts and Sciences said as it launched a crackdown on the use of AI.
- AI Actors And Writers Not Eligible For Oscars: Academy
AI Actors And Writers Not Eligible For Oscars: Academy Barron's
- AI Officially Can't Win an Oscar for Acting or Screenwriting
The new rules state that scripts must be “human-authored to be eligible” and acting nominations must be “demonstrably performed by humans with their consent.”
- Techno utopia or AI nightmare? The problem with music made by machines
Songs created by artificial intelligence are being met with disgust, but isn’t the whole point of electronic music that it’s made by robots?
- Spotify labels human artists with badge as AI content grows on platform
Spotify rolls out Verified badge and new artist details to improve transparency and help users identify authentic artists on its platform
- 'Verified By Spotify' is the music streamer's new way to help you avoid AI artists, but it doesn't go as far as I'd like — especially with Deezer, Qobuz and Apple Music making major anti-AI steps
The Verified By Spotify badge lets you find out if an artist is real on the streaming giant, but you still can't filter out AI slop in playlists
- Spotify unveils verified badge to distinguish humans from AI
Last year, Spotify removed more than 75m spam tracks from its platform. Read more: Spotify unveils verified badge to distinguish humans from AI
- Spotify's new 'Verified' badge confirms musicians like Taylor Swift, Drake aren't AI
Spotify's new 'Verified' badge confirms musicians like Taylor Swift, Drake aren't AI CBC
- Spotify Cracks Down on AI Content With New Feature. Here’s What to Look for
The new rollout is still in beta, but should soon be available worldwide.
- Spotify now verifies artists that are human, not AI
Spotify announces new verification tools to help listeners determine whether or not a profile or song is AI-generated, including badges and expanded artist information.
- Spotify cracks down on flood of AI music with verification badges
Spotify cracks down on flood of AI music with verification badges San Francisco Chronicle
- Taylor Swift vs. AI
Taylor Swift vs. AI marketplace.org
- US Navy turns to AI firm Domino for options to counter Iranian mines
US Navy turns to AI firm Domino for options to counter Iranian mines The Straits Times