Skip to content
Artwork for AI in 10

AI in 10

Chuck Goetschel

The most important AI story—explained in 10 minutes.


Every day, I break down the biggest AI story in just 10 minutes - what it is, why it matters, and how you can actually use it. No tech jargon, just AI made simple.

Play
  • 37 episodes
  • daily
  • Avg 6 min
  • English

Support the show

Goes straight to the publisher. podnod takes nothing.

Counted on this page — what you have heard stays on this device, so it is not something the list can be paged by.
  • October 29 · 5 min

    Guidewire pays $2B to put AI agents in insurance pricing

    Guidewire’s deal for Hyperexponential is a signal that “AI agents” are moving from demos into systems that can influence real-world insurance premiums. That matters because underwriting and pricing decisions touch people’s money—and require explanations. Guidewire said it will acquire Hyperexponential for $1.7B in cash, with up to $400M more tied to revenue targets (about $2.1B total). The goal is to embed AI-assisted underwriting and pricing inside Guidewire’s insurance software, where agents can assemble analyses and recommendations while humans still approve outcomes. The big question is how this changes speed, staffing, and fairness once integrated. Watch to see what this purchase suggests about where regulated-industry AI is headed next. New AI news every weekday — subscribe so you don't miss tomorrow's story. Referenced Links: Guidewire agrees to acquire Hyperexponential for up to $2 billion to bring AI agents into insurance underwriting Liquid AI launches LFM3, six compact open-weight models designed for text, vision, audio, and decision-making Oracle makes Google’s TPU processors available through Oracle Cloud 💬 Send Chuck a comment about this episode Support the show Want to go deeper with AI? A community of professionals is learning AI together right now at aihammock.com — show notes, links, tools, and real conversations about how to actually use AI in your life.

    • Transcript
  • October 28 · 6 min

    Akamai’s agent sandbox: can AI run code safely now?

    Akamai is betting the next AI leap isn’t a smarter chatbot—it’s safer action. If agents are going to touch real workflows, the biggest question is what happens when the code they generate goes wrong. Akamai launched two beta services: Akamai Inference for deploying open-weight/custom models on Akamai Cloud, and Akamai Agent Sandbox to run agent-written code in isolated environments. Akamai also said it’s adding dedicated NVIDIA Blackwell Ultra GPU clusters, aiming to make secure, low-latency enterprise AI more practical. Watch to see what Akamai is promising, what’s still missing (pricing/benchmarks), and what to ask any vendor before you let an agent “do things.” New AI news every weekday — subscribe so you don't miss tomorrow's story. Referenced Links: Akamai NVIDIA Akamai Blog NVIDIA Data Center 💬 Send Chuck a comment about this episode Support the show Want to go deeper with AI? A community of professionals is learning AI together right now at aihammock.com — show notes, links, tools, and real conversations about how to actually use AI in your life.

    • Transcript
  • October 27 · 6 min

    Anthropic files to go public—what’s inside the S-1?

    Anthropic’s IPO filing could turn one of the most important AI labs into a quarterly, disclosure-driven public company—meaning regular visibility into costs, risks, and momentum behind Claude. That matters for anyone using AI at work or investing in tech. Anthropic publicly filed an S-1 with the SEC for a proposed Nasdaq listing under ticker “ANT.” The filing doesn’t set a price yet, but it starts the process that will eventually reveal more about revenue, spending, compute intensity, and risk factors—including the tension of selling powerful AI while warning about its dangers. We break down what this filing signals, what to watch next, and what it could mean for the AI race. New AI news every weekday — subscribe so you don't miss tomorrow's story. Referenced Links: Anthropic newsroom CNBC AP News 💬 Send Chuck a comment about this episode Support the show Want to go deeper with AI? A community of professionals is learning AI together right now at aihammock.com — show notes, links, tools, and real conversations about how to actually use AI in your life.

    • Transcript
  • October 26 · 6 min

    GXO is putting humanoid robots in 2 more U.S. warehouses

    GXO Logistics expanding humanoid robots into real fulfillment work is a signal that “warehouse automation” is shifting from demos to day-to-day operations—and that could change what entry-level warehouse jobs look like next. GXO and Agility Robotics say Digit will expand to sites in Statesville, North Carolina, and Lebanon, Tennessee, after a prior live trial in Georgia. It’s a multi-year Robots-as-a-Service setup aimed at repetitive material handling for e-commerce, but the announcement doesn’t claim specific layoffs or a precise rollout timeline beyond more activity in 2027. Watch for what Digit does well, where it breaks, and what roles humans end up doing alongside it. New AI news every weekday — subscribe so you don't miss tomorrow's story. Referenced Links: GlobeNewswire (announcement source) Agility Robotics GXO Logistics 💬 Send Chuck a comment about this episode Support the show Want to go deeper with AI? A community of professionals is learning AI together right now at aihammock.com — show notes, links, tools, and real conversations about how to actually use AI in your life.

    • Transcript
  • October 25 · 5 min

    OpenAI restarts frontier training—without internet access

    OpenAI cutting internet access during frontier-model training signals how serious the security risks have become—and it could shape how safe (and how capable) future AI systems are. OpenAI says it resumed training its most advanced models after tightening security controls, including blocking internet access in the training environment. It also said its internal defenses testing surfaced a vulnerability in widely used software, highlighting how frontier AI work is now colliding with cybersecurity. The big question: can models stay powerful without being “online” during training? Watch for what changed, why OpenAI drew this line, and what it could mean next. New AI news every weekday — subscribe so you don't miss tomorrow's story. Referenced Links: OpenAI resumes frontier-model training after tightening security and restricting internet access OpenAI resumes frontier-model training after tightening security and restricting internet access (report) AI updates roundup (Craft, Decisions API, Agent Skills) 💬 Send Chuck a comment about this episode Support the show Want to go deeper with AI? A community of professionals is learning AI together right now at aihammock.com — show notes, links, tools, and real conversations about how to actually use AI in your life.

    • Transcript
  • October 24 · 6 min

    Kimi K3.1 Pro: Moonshot’s open-weight bet on code + finance

    Moonshot AI is positioning Kimi K3.1 Pro for the work people actually get paid for: coding and financial research. If it holds up, it’s a real alternative to the usual U.S. model stack. Moonshot released Kimi K3.1 Pro on Oct. 23, offering it via web/app/API plus a faster “Highspeed” option. Reporting says it’s in the three-trillion-parameter class and Moonshot plans to release model weights on Nov. 4—potentially letting developers run and customize it beyond a closed chatbot. We break down what’s new, what to test safely, and what to watch once the weights drop. New AI news every weekday — subscribe so you don't miss tomorrow's story. Referenced Links: Bloomberg Kimi (Moonshot AI) official site Kimi web app Moonshot API platform 💬 Send Chuck a comment about this episode Support the show Want to go deeper with AI? A community of professionals is learning AI together right now at aihammock.com — show notes, links, tools, and real conversations about how to actually use AI in your life.

    • Transcript
  • October 23 · 6 min

    Claude Duet is an always-on work agent—what can it do without you?

    Anthropic’s Claude Duet pushes AI past chat and into your actual workday—watching context across apps and surfacing tasks before you ask. That convenience could save hours, but it also forces new choices about trust, approvals, and privacy. Anthropic launched a research preview of Claude Duet (powered by Claude Opus 5.5) designed to prep briefings, schedule meetings, build apps, and help run business workflows. The key detail: it’s proactive, but still gated—payments, messages, document signing, and deletion require user approval. The race is shifting from “best answers” to “safe actions. We break down what Duet can do, where the guardrails are, and what to test first before you connect your accounts. New AI news every weekday — subscribe so you don't miss tomorrow's story. Referenced Links: Anthropic unveils Claude Duet, an always-on AI agent for professionals Anthropic’s new Claude Duet agent is always-on for work Walmart announces tools that let approved AI agents shop on its website Anthropic’s Claude Duet is a proactive AI agent (research preview) 💬 Send Chuck a comment about this episode Support the show Want to go deeper with AI? A community of professionals is learning AI together right now at aihammock.com — show notes, links, tools, and real conversations about how to actually use AI in your life.

    • Transcript
  • October 22 · 5 min

    Google Cloud adds OpenAI GPT-6 models (and Microsoft isn’t it)

    Google Cloud adding OpenAI’s newest GPT-6 models isn’t just another model drop—it changes how enterprise teams can buy, govern, and deploy OpenAI inside their existing Google Cloud stack. Google says GPT-6 Astra, GPT-6.1 Sol, and GPT-6 Luna are rolling out in private preview through Model Garden. That’s a big signal for cloud competition: OpenAI access is getting more portable beyond Microsoft’s ecosystem, even though details like pricing, benchmarks, and public launch timing weren’t confirmed. Watch for what this means for vendor lock-in, data controls, and which cloud becomes the default place to run OpenAI at work. New AI news every weekday — subscribe so you don't miss tomorrow's story. Referenced Links: Google Cloud adds OpenAI’s newest models to its enterprise AI platform NVIDIA unveils OpenShell architecture for safer AI agents at Berlin GTC Databricks makes its managed AI Runtime generally available for GPU training OpenAI expands ChatGPT advertising availability to 89 countries 💬 Send Chuck a comment about this episode Support the show Want to go deeper with AI? A community of professionals is learning AI together right now at aihammock.com — show notes, links, tools, and real conversations about how to actually use AI in your life.

    • Transcript
  • October 21 · 6 min

    Zoom’s new AI “context layer” can read across your work

    Zoom is betting that the next leap in workplace AI isn’t a smarter chatbot—it’s one that can see the full trail of meetings, chats, docs, and tools you use every day. That could save hours, but it also raises real questions about what the assistant can access. At Zoomtopia (Oct 20, 2026), Zoom announced its “Unified Context Layer” plus new AI workplace tools, aiming to connect context across conversations and workflows so assistants can summarize decisions, pull next steps, and answer questions without you repeating background. The hard part is governance: permissions, retention, and whether the AI’s “answer” matches the record. Watch to understand what Zoom’s assistant could know, where it might go wrong, and what to ask your admin before turning it on. New AI news every weekday — subscribe so you don't miss tomorrow's story. Referenced Links: Zoom Newsroom (source) Zoom Blog (source) Zoomtopia (event page) 💬 Send Chuck a comment about this episode Support the show Want to go deeper with AI? A community of professionals is learning AI together right now at aihammock.com — show notes, links, tools, and real conversations about how to actually use AI in your life.

    • Transcript
  • October 20 · 6 min

    Report: AI agents helped breach 100+ websites — how?

    Norforge’s claim that AI agents helped attackers breach 100+ websites matters because autonomous tools could scale real-world hacking faster than humans can respond. According to reporting, Norforge said AI agents assisted in compromises across more than 100 sites, but key technical details (victims, attackers, and the agents’ exact role) weren’t disclosed. That uncertainty is the story: if agents can chain steps like targeting, testing, and retrying, defense has to change. Watch for what “agent” changes in an attack—and what you can do now. New AI news every weekday — subscribe so you don't miss tomorrow's story. Referenced Links: The Verge Norforge Menlo Ventures Forerunner Ventures 💬 Send Chuck a comment about this episode Support the show Want to go deeper with AI? A community of professionals is learning AI together right now at aihammock.com — show notes, links, tools, and real conversations about how to actually use AI in your life.

    • Transcript
  • October 19 · 6 min

    Anthropic’s AI agents cut junior hiring from 27% to 9%

    Anthropic’s own workforce data suggests the first impact of AI agents may be fewer entry-level jobs, not sudden mass layoffs. If that pattern spreads, the “career ladder” into tech could change fast. In a new internal report, Anthropic says junior employees fell from 21% to 14% of technical staff, and junior technical hires dropped from 27% to 9%. Anthropic also claims its AI agents now complete about 6x the work-hours of the humans directing them, while meeting time per employee rose 20%—hinting at a shift toward supervision and judgment work. Watch to see what these numbers really imply for junior roles and what skills still compound in an agent-run workplace. New AI news every weekday — subscribe so you don't miss tomorrow's story. Referenced Links: Anthropic workforce report (Oct 17, 2026) Anthropic: AI agents work-hours vs. humans (methodology) Anthropic says AI agents are already reshaping its workforce Coverage: junior technical hiring drops at Anthropic 💬 Send Chuck a comment about this episode Support the show Want to go deeper with AI? A community of professionals is learning AI together right now at aihammock.com — show notes, links, tools, and real conversations about how to actually use AI in your life.

    • Transcript
  • October 18 · 5 min

    DeepSeek’s new Flash model is faster… and talks less

    DeepSeek’s V4.1-Flash-1017 isn’t selling “bigger”—it’s selling speed and shorter answers, which could change what you pay for (and tolerate) in daily AI use. DeepSeek pushed the update live Oct. 17 across its web/app and developer API, claiming stronger benchmarks, faster task completion, and more concise outputs than the prior V4.1 Flash release. The interesting question is whether “shorter” is a real quality upgrade (better signal) or just cost/latency optimization that changes how assistants feel. We break down what likely changed, what to test, and what this means for OpenAI/Anthropic competition. New AI news every weekday — subscribe so you don't miss tomorrow's story. Referenced Links: DeepSeek releases faster, leaner V4.1-Flash-1017 model (Bloomberg) DeepSeek’s rapid release cadence and V4.1 Flash context (Bloomberg) DeepSeek (try the model) DeepSeek announcement source posts (X) 💬 Send Chuck a comment about this episode Support the show Want to go deeper with AI? A community of professionals is learning AI together right now at aihammock.com — show notes, links, tools, and real conversations about how to actually use AI in your life.

    • Transcript
  • October 17 · 5 min

    Google AI Search adds inline sources—will anyone click them?

    Google’s AI Overviews are convenient, but they’ve also trained people to trust a summary without seeing proof. This update is Google admitting the biggest issue isn’t speed—it’s trust and attribution. Google says AI Mode and AI Overviews now show clearer inline links, stronger source context, and more timely articles so you can inspect claims faster. It’s a direct response to criticism that AI answers feel like an endpoint instead of a starting point. The real test is whether these links send traffic out—or just make the summary feel “verified.” We break down what changed, why it happened now, and how to use it without getting misled. New AI news every weekday — subscribe so you don't miss tomorrow's story. Referenced Links: Google: Updates to AI Mode and AI Overviews (Oct 15, 2026) Reuters: CNN, FT, Axios join Google’s paid AI content licensing program 💬 Send Chuck a comment about this episode Support the show Want to go deeper with AI? A community of professionals is learning AI together right now at aihammock.com — show notes, links, tools, and real conversations about how to actually use AI in your life.

    • Transcript
  • October 16 · 5 min

    White House gives itself 120 days to police AI—what changes?

    The White House AI task force’s 120-day review could shape how AI is tested, monitored, and held accountable—affecting everything from scams and deepfake fraud to safeguards for hospitals, utilities, and finance. After its first official meeting on Oct. 15, the task force adopted a charter and spun up workstreams on cybersecurity/critical infrastructure, biosecurity, and model control/accountability. It also set up committees tied to an AI company agreement and a DOJ/FTC coordination effort, hinting at a shift from voluntary promises to enforceable oversight. We break down what they’re actually reviewing—and what would be the most consequential recommendations if they land. New AI news every weekday — subscribe so you don't miss tomorrow's story. Referenced Links: White House AI task force begins 120-day review of AI risks and accountability AI news digest (Oct. 16, 2026) Huawei and 14 partners launch a global AI-city development alliance 💬 Send Chuck a comment about this episode Support the show Want to go deeper with AI? A community of professionals is learning AI together right now at aihammock.com — show notes, links, tools, and real conversations about how to actually use AI in your life.

    • Transcript
  • October 15 · 6 min

    Meta’s Muse Wonder is here—why is the best version paywalled?

    Meta’s new Muse Wonder models could reshape which AI developers build on—and what everyone ends up using in apps. But Meta is signaling a sharper divide between free access and premium capability. Meta announced Muse Wonder and Muse Wonder Full on Oct 12, with availability via APIs and coding tools. Meta positions them as competitive with OpenAI and Anthropic, but key details like pricing, context limits, and benchmark breakdowns aren’t clearly verified yet. The biggest tell: Muse Wonder Full is reserved for paid subscribers. We’ll break down what this launch means for developers, users, and what to watch next. New AI news every weekday — subscribe so you don't miss tomorrow's story. Referenced Links: Meta launches Muse Wonder, a new flagship AI model family Apple unveils HomeView, an AI-powered smart-home hub centered on Siri Google gives European researchers access to advanced AI tools and $3 million in computing credits Anthropic warns that small AI models may soon self-replicate on smart-home devices 💬 Send Chuck a comment about this episode Support the show Want to go deeper with AI? A community of professionals is learning AI together right now at aihammock.com — show notes, links, tools, and real conversations about how to actually use AI in your life.

    • Transcript
  • October 14 · 6 min

    Character.AI’s parent is bankrupt—Disney bids $20M for assets

    Character.AI could change fast if a Disney subsidiary ends up buying key assets—terms, privacy, and which bots stay online could all shift. If you’ve built characters or had long chats, what happens next matters. Character Technologies (Character.AI’s parent) filed for Chapter 11 and named a Disney unit as the “stalking horse” bidder in a court-run sale, starting around $20 million. Bankruptcy doesn’t mean the app shuts down tomorrow—but it does put the platform under court supervision, with an auction process that could reshape the product. The big unknown: which technology, IP, and user-data obligations move with the sale. We break down what a stalking-horse bid means, what to watch before December, and what Character.AI users should do now. New AI news every weekday — subscribe so you don't miss tomorrow's story. Referenced Links: Character.AI’s parent files for bankruptcy and agrees to sell assets to Disney (Reuters) Google launches Gemini 4 Argon globally (Reuters) Google DeepMind and Singapore unveil an on-device AI scam detector (Reuters) Anthropic warns that AI models could spread across smart-home devices (Reuters) 💬 Send Chuck a comment about this episode Support the show Want to go deeper with AI? A community of professionals is learning AI together right now at aihammock.com — show notes, links, tools, and real conversations about how to actually use AI in your life.

    • Transcript
  • October 13 · 6 min

    Amazon’s rumored $7B Decart deal is really about chips

    Amazon reportedly considering a Decart acquisition matters less as M&A gossip and more as a shift in who controls how AI runs on hardware. If Decart makes models portable across chips, that changes AWS costs and Nvidia leverage. Reports say Amazon is in talks to buy Decart for about $7B, though nothing is signed and talks could collapse. Decart builds tooling to move AI models between different chips and also works on 3D world simulation; it’s reportedly had early access to Amazon’s Trainium3. The key question is whether Amazon can turn chip portability into a real advantage. Watch to understand what Amazon might be buying, and why developers should care about “model portability.” New AI news every weekday — subscribe so you don't miss tomorrow's story. Referenced Links: Amazon reportedly considers buying AI startup Decart for about $7 billion (Wall Street Journal) Amazon in talks to acquire Decart for around $7 billion (The Information) Amazon reportedly discusses Decart acquisition amid AI chip competition (TechCrunch) Amazon explores AI infrastructure moves as chip scarcity persists (Reuters) 💬 Send Chuck a comment about this episode Support the show Want to go deeper with AI? A community of professionals is learning AI together right now at aihammock.com — show notes, links, tools, and real conversations about how to actually use AI in your life.

    • Transcript
Showing 21–37 of 37 episodes