DX Today | No-Hype Podcast & News About AI & DX
The DX Today Podcast: Real Insights About AI and Digital Transformation
Tired of AI hype and transformation snake oil? This isn't another sales pitch disguised as expertise. Join a 30+ year tech veteran and Chief AI Officer who's built $1.2 billion in real solutions—and has the battle scars to prove it.
No vendor agenda. No sponsored content. Just unfiltered insights about what actually works in AI and digital transformation, what spectacularly fails, and why most "expert" advice misses the mark.
If you're looking for honest perspectives from someone who's been in the trenches since before "digital transformation" was a buzzword, you've found your show. Real problems, real solutions, real talk.
For executives, practitioners, and anyone who wants the truth about technology without the sales pitch.
DX Today | No-Hype Podcast & News About AI & DX
DX Today AI Daily Brief - Saturday, March 28, 2026
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
It's Saturday, March 28, 2026. You're listening to the DX Today AI Daily Brief. Today, cybersecurity stocks crash after Anthropic's leaked AI model raises fears of an offensive cyber arms race. OpenAI's ad business hits$100 million in just six weeks. And Wikipedia draws a hard line against AI-generated content. Let's get into it.
SPEAKER_01The fallout from Anthropic's Claude Mythos leak sent shockwaves through financial markets on Friday. Cybersecurity stocks plunged after the leaked internal documents described the model as far ahead of any other AI in offensive cyber capabilities. CrowdStrike and Palo Alto networks each fell more than 7%. Zscaler and Sentinel-1 dropped over 8%, while Tenable lost nearly 11%. The Global X Cybersecurity ETF closed at its lowest level since November 2023, extending its decline this year to more than 21%. Investors fear that a new generation of AI models, capable of finding and exploiting software vulnerabilities at scale, could fundamentally undermine the defensive tools that the entire cybersecurity industry is built upon.
SPEAKER_00OpenAI's advertising pilot for ChatGPT has crossed$100 million in annualized revenue in less than six weeks since launching in the United States. More than 600 advertisers have signed onto the program, which places clearly labeled ads below ChatGPT responses without influencing the chatbot's answers. About 85% of free and go plan users are eligible to see ads, though fewer than 20% are shown them on any given day. Users under 18 are excluded, and ads will not appear near content related to politics, health, or mental health. Nearly 80% of participating small and medium-sized businesses have signaled interest in continuing. OpenAI plans to expand the program to Canada, Australia, and New Zealand, with a self-serve advertiser portal expected in April.
SPEAKER_05A$10 billion bet on AI infrastructure.
SPEAKER_04Meta has increased its investment in a new data center in El Paso, Texas, from$1.5 billion to$10 billion, a more than six-fold jump. The facility is designed to deliver one gigawatt of AI compute capacity when it opens in 2028. The expansion reflects Meta's aggressive push to build infrastructure for its growing AI workloads, which now span everything from recommendation algorithms to the company's family of Llama models. Meta says the project will prioritize renewable energy and water efficiency and create thousands of construction and permanent jobs. The announcement comes as the company is simultaneously facing scrutiny over plans to lay off as many as 16,000 employees, roughly 20% of its workforce, as it redirects resources toward AI development. Microsoft Moves In Next Door.
SPEAKER_02Microsoft is stepping in to build two new AI factory buildings and an on-site power plant on a tract of land in Abilene, Texas, that OpenAI decided not to expand further. Data center developer Crusoe announced the 900-megawatt campus on Friday, bringing the total capacity at the site to approximately 2.1 gigawatts. The new Microsoft facilities will sit directly adjacent to the complex Crusoe has been building for OpenAI and Oracle as part of the Stargate initiative, a project so large that President Trump personally announced it after his inauguration. Microsoft was once OpenAI's exclusive cloud provider and still holds a 27% stake, but the two companies are increasingly pursuing separate AI infrastructure strategies, even as they remain literal neighbors in the Texas mesquite.
SPEAKER_05Wikipedia draws the line on AI.
SPEAKER_03Wikipedia has officially banned the use of AI-generated text by its editors, updating its guidelines to state that the use of large language models to generate or rewrite article content is prohibited. The policy clarifies earlier, vaguer language that merely discouraged creating articles from scratch with AI. Editors may still use AI for basic copy editing that does not add new content, and for translating articles from other language versions into English, provided they follow existing translation rules. The ban reflects growing concerns about AI-produced text, introducing factual errors, hallucinated citations, and stylistic inconsistencies into the world's largest open encyclopedia. Wikipedia has long relied on a volunteer editing model built on verifiability and sourcing, and the new rules aim to protect that foundation from what critics have called AI-generated slop.
SPEAKER_05ByteDance fills the video AI gap.
SPEAKER_01ByteDance has launched Dreamin' a CDance 2.0, its most powerful AI video generation model, directly inside its editing platform, CapCut. The rollout includes a new feature called Video Studio, a timeline-free workspace where AI handles scripting, character development, storyboarding, footage generation, and final assembly in a single flow. At launch, the model supports clips up to 15 seconds across six aspect ratios. The initial release targets users in Southeast Asia, the Middle East, Latin America, and Africa, with broader global availability coming later. The timing is notable. OpenAI shut down its competing Sora video app earlier this week, and ByteDance appears to be moving quickly to capture the gap. The model will also power ByteDance's DreamA Generation platform and its Pippet marketing tool.
SPEAKER_05A new personal AI challenger emerges.
SPEAKER_00Backed by$100 million of Adcock's own money, the company is building multimodal AI systems, custom hardware, and user interfaces under one roof. Hark systems are designed to see, listen, and interact in real time while maintaining a persistent memory of the user's life. The 45-person team includes former researchers from Meta's AI division and designers from Apple and Tesla. Adcock says the first AI models will ship this summer, followed by a family of dedicated hardware devices that will be distinct from existing phones, wearables, and smart glasses. The goal, he says, is for the products to feel so essential that going without them would be like a day of lost information.
SPEAKER_05Enterprise AI gets a voice.
SPEAKER_04Eleven Labs in IBM have announced a partnership to bring premium voice capabilities to enterprise AI agents. The integration adds 11 labs text-to-speech and speech-to-text technology to IBM's Watson X Orchestrate platform, giving businesses access to more than 10,000 voices across 70 languages. The collaboration includes enterprise grade security features such as PCI compliance for payment processing, a zero retention mode designed for HIPAA compliant data handling, and regional data residency options. The partnership positions Watson X as a platform where companies can build voice-enabled AI agents that handle customer interactions, internal workflows, and multilingual communications at scale. The announcement reflects a broader industry shift toward voice as a primary interface for agentic AI systems in business environments. OpenAI opens a new bounty program.
SPEAKER_02OpenAI has launched a safety bug bounty program designed to identify AI abuse and safety risks that go beyond traditional security vulnerabilities. Hosted on the Bug Crowd platform, the program covers agentic risks including model context protocol abuse, third-party prompt injection, and data exfiltration. It also targets violations of account and platform integrity, such as bypassing anti-automation controls or manipulating trust signals. The program complements OpenAI's existing security bounty, which has rewarded 409 vulnerabilities since 2023. Notably, general jailbreaks that only produce rude language or easily searchable information are out of scope. OpenAI says it is looking for flaws that enable direct user harm with actionable fixes, reflecting the company's growing focus on real-world safety as its agents take on increasingly complex tasks.
SPEAKER_05China's latest AI contender.
SPEAKER_03Xiaomi has released MIMO V2 Pro, a trillion-parameter large language model that the company says ranks eighth globally on the Artificial Analysis Intelligence Index. The model uses 42 billion active parameters with a hybrid attention mechanism and supports a 1 million token context window. Xiaomi says coding performance surpasses Clawed 4.6 Sonnet, while general agent capabilities approach Opus 4.6 on major benchmarks. MyMO V2 Pro is designed as a native brain for the OpenClaw Agent framework, with fine-tuning across complex agent scaffolds for stronger tool call accuracy and multi-step reasoning. The model's API is now publicly available with tiered pricing starting at$1 per million input tokens. Xiaomi says the release represents a milestone in its pursuit of artificial general intelligence.
SPEAKER_05Meta acquires an AI startup.
SPEAKER_01Meta has acquired Dreamer, an AI agent startup, and brought its founding team into the company's Superintelligence Labs Research Division. The acquisition, reported by the information, is part of Meta's broader push to strengthen its personal AI capabilities as competition with OpenAI and Anthropic intensifies. Dreamer's technology focuses on autonomous agents that can plan and execute multi-step tasks, an area that has become a critical battleground for major AI labs. The deal comes alongside Meta's massive infrastructure investments, including the$10 billion El Paso data center and its ongoing development of custom MTIA chips and the newly announced ARM AGI CPU partnership. Financial terms of the Dreamer acquisition were not disclosed, but the move signals Meta's intent to compete not just in foundation models, but in the agent platforms built on top of them.
SPEAKER_05And finally, the AI layoff toll mounts.
SPEAKER_00The wave of AI-driven job cuts across the technology sector continues to grow. Oracle is reportedly planning to eliminate between 20,000 and 30,000 positions as it redirects spending toward AI data centers. Meta is moving forward with plans to cut roughly 16,000 workers, about 20% of its workforce. These follow earlier cuts at Block, which reduced staff by 40%, and at Lassian, which cut 10%. Combined with the 55,000 AI attributed layoffs in 2025, the trend line is steepening. A recent Duke CFO survey estimated 502,000 AI-related job losses across the US economy this year. The layoffs are concentrated in roles that AI agents and automation tools can now perform, particularly in white collar functions like customer service, content moderation, and routine engineering tasks.