AI

2026-05-25 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-24 · open the original ↗

• Hallucinations usually signal missing context, not a faulty model. The guests' book argues that hallucinations are gaps in "canon", the information a system accepts as authoritative. The guest answering says that with 2026 models, a model told you live on the third floor won't say the second. A hallucination therefore flags that the context is missing or that retrieval failed to fetch it. The host adds an example of his own: an HR agent given an undated two-year-old guide whose policy has since reversed, which would answer confidently and wrongly.
• A "canon manager" should own what agents treat as true. The authors coined the title to consolidate roles that organizations name differently. The anti-pattern they describe is critical agent context buried in a GitHub repo that only an engineer can change, when the engineer is not the subject matter expert. The canon should be moved in front of the experts who are accountable for the agent's behaviour.
• AI tools that cut friction make company wikis viable. The guest says no one used the old company wiki, but a well-maintained one now unlocks "a million things". The biggest gain for their company and others has been AI that spots conflicts in the knowledge base and proposes a diff for an expert to accept or tweak.

2026-05-23 · open the original ↗

• AI treats whatever data it is given as truth. So, Mumford argues, AI engineering falls back on data engineering. The book quotes Anchorman: whatever you put on the teleprompter, Ron Burgundy will read. Mumford likens this to data scientists who are hired, find the company's data in a mess, and spend two years doing data engineering. The hope is to pass that lesson on to AI engineering.
• AI engineering's data problem is unstructured documents, not missing numbers. Data science asked where the numbers were. Mumford says AI engineers in 2026 ask where the PDFs, images and Google Docs are, which are authoritative and which are junk, and how to chunk, embed and search them.

2026-05-22 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-22 · open the original ↗

• Recent computer science grads now fare worse than the average new graduate. The host cites the New York Fed: 7% unemployment for computer science grads aged 22 to 27 and nearly 8% for computer engineering. That compares with about 6% for all recent grads and about 4% for the US workforce. He adds that US computer science enrollment fell 11% in 2025 and programming enrollment 26%, and that Handshake postings are down roughly 50% from their 2022 peak. Revelio Labs data suggest entry-level software and data analytics postings fell by as much as 67%. He contrasts this with White House adviser Kevin Hassett's claim of no sign in the data that AI is costing jobs.
• Some evidence says AI is shutting the young out of entry-level jobs. A 2025 study by Stanford's Erik Brynjolfsson and colleagues found employment among 22 to 25-year-olds in AI-exposed jobs down 13% since 2022, while older and less-exposed workers held steady or rose. The host says the Dallas Fed replicated the pattern this year and found it driven by non-hiring, not layoffs.
• Other studies find no AI effect, blaming rates and over-hiring instead. Two Google economists found posting declines in AI-exposed occupations as steep for seniors as juniors, and starting before ChatGPT. A Federal Reserve study of over a million firms found "precisely estimated null effects". Skeptics point to high interest rates, the end of the tech boom and 2021 over-hiring. Stanford's Eric Roberts likens the moment to the dot-com bust, after which hiring recovered by 2004.
• The host's view: macro explains much, but entry-level exclusion is real. He says the granular evidence that young AI-exposed workers are uniquely shut out is hard to dismiss. At his own consulting firm, Claude Code and OpenAI's Codex make code so cheap and fast that he is not sure why it would hire any entry-level developer in the foreseeable future.
• Grads should differentiate on design, domain, portfolio, agents and network. The host's five tips: compete on system design rather than raw coding output; pair AI fluency with a domain such as healthcare or law; build a public portfolio, since recruiters distrust CVs and applicant-tracking systems filter most out; learn agentic tooling such as RAG, evaluation and multi-agent orchestration, citing PwC's finding of a 56% wage premium for AI skills; and rely on referrals over job boards. He says the advice applies at any career stage.

2026-05-22 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-21 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-21 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-21 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-21 · open the original ↗

• Zillow Offers lost to a winner's curse once its features drifted. The host gives the loss as almost a billion dollars. Miller describes the trap: an offer set too high is accepted and Zillow loses, while one too low is refused. The pilot went well at first, but COVID and other changes caused model drift. Miller treats any adversarial position as a shift in the feature space the model never planned for.
• Modern AI products face far more adversarial environments. Miller would hate to build a customer-service bot that hands out discounts to settle complaints, since Reddit posts teach people to trick it. Such projects need observability that flags out-of-norm behaviour immediately. At the start of COVID, Miller says, many data science teams threw out their models and forecast on the fly because the present no longer resembled the past.

2026-05-21 · open the original ↗

• Define and size the move before slicing anything. The presenter, a co-host, calls this "level zero": state the metric's definition, grain and window, say whether the change is absolute or relative and against what (prior four-week average, week-over-week, year-over-year), and set a bar for whether it is worth chasing. A 12% drop that follows an unexplained 12% rise is just a return to baseline.
• Drill into the dimension that explains the most change. The biggest percentage move is not the target. In the presenter's example, revenue is down 12% ($5 million) and APAC accounts for $4.5 million of it. That is the direction to follow, even if one acquisition channel is down 80%. Slice again inside that segment while it still carries most of the top-line move, from "EMEA" down to paid members in Germany or the UK on iOS. The failures named are panic, telling a story from the one dimension a dashboard has, and "drilling to nowhere", which Shane adds turns up spurious correlations and red herrings.
• Agents add depth a human would skip, not just speed. Shane recalls spending two months at Nextdoor on a root cause analysis of falling sessions, losing track of threads as new data came in. He argues that subagents can drill dimensions in parallel and write to one knowledge base, and that even naive segmentation and attribution done this way is "extremely powerful", with causal-inference methods as an optional layer on top.
• The demo traced a June 2024 ticket spike to one iOS release. On the repo's fictional Novamart data (50,000 users, 47,000 orders, 6.5 million events, 500 products, a $99 membership), a plain-English prompt had Claude Code find that payment tickets jumped to 41% of all tickets in June and fell back in July while attempts grew smoothly. It named a defective iOS app release, version 2.3.0, fixed a few days later, and rated its own confidence high. The presenter says the work still has to be checked. The Google Doc shown was produced beforehand, and the presenter says they already knew the story the data held.
• Opus is required for reasoning work, by the hosts' account. Shane says the system "does not work well with Sonnet", which is "very confidently wrong". It only clicked with Opus 4.6 in February, and Opus 4.7 went "off the rails" until the paid version of the repo was updated. A third co-founder says Sonnet can do pure execution of fixed skill rules. In Q&A the presenter says the agent flags every cause it finds when several drive a move, not only the largest. It connects to Snowflake and DuckDB natively, has no Copilot version, and a Codex version is being built.

2026-05-20 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-20 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-20 · open the original ↗

• With near-instant prototypes, an AI project must prove value on day one. The guest says enterprises hold on to ideas too long because traditional software rewarded sticking with a partial product. Their example is the text-to-SQL chatbot that, they say, every company tried, including theirs early on at Pattern: a day yields an impressive demo, so companies commit and keep going. The guest cites AWS, which had just announced that six engineers rebuilt its AI infrastructure over two or three months in a move from Bedrock to "Project Mantle", as the resolve to start over.
• Once building stops being the bottleneck, definition and feedback decide outcomes. In the define-build-feedback cycle, slow building used to hide weak definition and feedback. The organizational fix the guests describe is narrow lanes: two or three people with a clear target and freedom to decide, who fail more readily because they are not failing in front of a large audience.

2026-05-19 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-19 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-19 · open the original ↗

• Design every process so a human, an AI or both can run it. The book's "user agnosticism tenet" is illustrated by a sales chatbot. A human can join at any point and pick up where the bot left off, so automating the pipeline does not lock the human out and a later rollback does not force a massive refactor.
• When building is cheap, definition and feedback become the bottleneck. Prototypes are near-instant, so the guests say AI projects must prove value on day one or be scrapped. Their example is the text-to-SQL chatbot every company tried, whose day-one demos lured teams into persisting. Miller's fix is narrow lanes of two or three people who can fail repeatedly without a large audience. A guest cites AWS's just-announced rebuild of its AI stack by six engineers over two or three months as the resolve needed.
• Build workflows before agents, and remove AI where a rule will do. If you aren't happy with a workflow where you set the path, the guests ask, why trust an agent improvising one? They see teams progress from skills (markdown files subject matter experts can iterate on) to workflows to agents. Many "agent builders" are really workflow builders, they say, and deterministic steps cut cost and risk. One guest swapped an AI step for a decision tree of canned statements and user complaints vanished overnight.
• Hallucinations mostly mean the canon is missing or wrongly retrieved. Mumford invokes Anchorman (Ron Burgundy will read whatever is on the teleprompter) to argue that AI engineering is now mostly data engineering over unstructured PDFs and docs. The guests propose a "canon manager" who owns authoritative context. That context should move out of engineers' GitHub repos to accountable subject matter experts, with AI tools proposing diffs to resolve conflicts.
• Fine-tune rarely, and treat speed as the only durable moat. A guest says more people fine-tune than should; the case is a narrow, stable task such as classification at millions or billions of data points. Miller weighs a gain of 93 to 95% accuracy against the projects forgone, and the host cites BloombergGPT's eight-figure cost and the few weeks before GPT-4 eclipsed it. Since everyone has the same frontier models, Mumford argues advantage comes from cycling define-build-feedback faster, with taste setting how much each cycle improves.

2026-05-19 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-18 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-18 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-18 · open the original ↗

• Visual checks at every pipeline step catch errors early and cheaply. Manz calls himself a "visualization person" who wants to see the data at each stage. He says he has often loaded a dataset with problems such as missing values that a quick visual check would have caught. Missing them cost him an afternoon down the wrong path, or baked the flaw into a result.
• Scientific visualization tools should live where the data work happens. In biomedical fields, Manz says, new experimental techniques yield abstract data with no familiar representation, so tools such as the genome browser had to be invented. Microscopy tools tend to be one-off applications outside the analysis environment. He argues for composable tools at hand for cheap checks, such as overlaying a segmentation on a tissue image to test the thresholding.

2026-05-18 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-17 · open the original ↗

• Coding agents are already a form of recursive language model. Manz describes recursive language models, from research released last year, as a model with an environment it can offload context to rather than stuffing everything into the prompt. He says this counters "context rot", where models forget or go off track as context grows. For today's coding agents that environment is the file system, with tools to fetch context when needed.
• Marimo pair extends the agent's environment into live Python memory. With Marimo pair, Manz says, context can be values held in memory as well as files, so the agent need not pull all the data's context into its window. The notebook becomes a structured representation of the agent's working memory. Ask a plain coding agent the same data question repeatedly and it might find 10 different ways through the file system; a notebook yields a similar scaffold each time, with each cell recording a step to the answer.

2026-05-16 · open the original ↗

• Data work now starts in the coding agent, which drives the notebook. Manz describes asking Claude Code, or open code or Codex, to start a Marimo pair session, then declaring at a high level which datasets and libraries to use. The skill teaches the agent to start or connect to Marimo. The agent then edits the notebook open in a browser while the human can still edit cells by hand.
• Shared kernel state lets the agent see what the human sees. Because the agent can run code in the kernel, read cells and take screenshots of them, a user can circle something interesting and ask about it. Manz says notebook state that used to be trapped becomes context beyond the file system. Streaming the human's edits back to the agent is not built yet; he mentions exploring MCP's "channels" feature.
• Agents make exploratory analyses cheap enough to try. Just as agentic coding made software ideas feel cheap, Manz says he can now ask for, say, five validations he would otherwise skip for not remembering an API. He then lets the agent make plots and return when he can decide.

2026-05-16 · open the original ↗

• Every success metric has a shadow that worsens under optimization. The presenter, a co-host, opens with a checkout feature that lifts conversion 15% while net revenue falls two weeks later. The pairs given are conversion rate against average order value, sign-ups against day-30 retention, engagement against NPS or complaints, and support speed against tickets reopened. The presenter argues that learning how metrics move against each other matters more than any single launch decision.
• Fix three things before launch: metric, guardrails, thresholds. The success metric must be specific ("purchase rate for new users", not "engagement"). Guardrails are named in advance rather than "track everything". Hunting for supporting metrics after launch is "data theater". Each guardrail gets an alert threshold (a 3–5% move might be tolerable) and an absolute kill threshold that pauses the feature.
• Pre-set thresholds keep the decision out of politics. Shane argues that fixing numbers in a document before any data exists keeps finance, sales and product teams from reading results through their own priorities. Without that, he says, post-launch conversations turn into conflict and get political. The presenter adds that the more sensitive the experiment, the more this matters.
• The demo designed guardrails for a "save for later" launch. On the fictional Novamart data, with a 30-day purchase rate target of +15%, Claude Code on Opus 4.7 flagged cannibalization of immediate purchases as the main risk. It proposed same-session purchase rate, average order value, return rate and 90-day repeat purchase rate. Told return rate isn't tracked, it substituted support-ticket rate and pre-shipment cancellation rate. For returns rising while 90-day purchases stayed flat, it read the feature as changing what customers buy, not whether they return. The presenter says the acceptable ranges came from earlier runs on the same data and can be set explicitly.
• Have a fresh-context subagent check the work, not the builder. Answering an attendee, Shane says the agent that did the work is biased toward judging it correct. He uses subagents with clean context and a skill in which several personas critique separately, argue, and converge. He also uses a second frontier model such as Codex. On data, he says to feed raw numbers, because the LLM writes Python to compute and never ingests them. Numeric work should stay in code to avoid hallucination, though rounding may help if the model pattern-matches on numbers directly.

2026-05-15 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-15 · open the original ↗

• AI compute demand is outrunning a supply chain with several simultaneous bottlenecks. The host dismisses "tokenmaxxing", maximizing tokens consumed in Claude Code or Codex to look good on internal usage dashboards, as a vanity metric. He cites the symptoms of scarcity: Anthropic's weekly Claude Code rate limits from last August, which it blamed on power users running it 24/7 and said it was "very compute constrained", OpenAI pausing Sora sign-ups, GitHub halting new Copilot subscribers, and unreliable on-demand H100s without reserved capacity. Fixing one bottleneck, he says, does not unstick the system.
• Packaging and memory, not GPU fabrication, are the chip choke points. He reports Nvidia data-centre GPU lead times of 36 to 52 weeks, Blackwell largely sold out through mid-2026, and H100 spot rents up about 30% since November. TSMC's CEO C.C. Wei has said CoWoS packaging is sold out through 2026; Nvidia reportedly holds about 60% of it through 2027, and new fabs take two to three years. HBM demand has roughly quintupled since 2023 (an H100 needs 80 GB, a B200 192 GB), and SK Hynix, Samsung and Micron are largely sold out into 2026. Nvidia reportedly cut RTX 50 output 30% to 40% as a result, and new HBM fabs take 18 to 24 months.
• Agentic AI needs far more CPUs per GPU. Morgan Stanley analysis cited by the host puts chatbot systems at about one CPU per 12 GPUs and agentic systems at one-to-one. Intel's CFO David Zinsner said on the Q1 call that its server CPU shortfall "starts with a B". The host says server CPU prices have risen 10 to 20% in a couple of months, a big part of why Intel's market cap more than doubled in 6 months.
• Electricity is the most intractable bottleneck. Hyperscalers report build-outs gated by grid interconnects (18 to 36 months) and transformers (another 18 to 24). Gartner projects power shortages will restrict 40% of AI data centres by 2027. The host counts at least 12 states with moratorium bills as of March, a Maine statewide moratorium passed in April and vetoed, and more than 50 local moratoria. He also cites rate rises of 8 to 15% near clusters in Virginia, Texas and Georgia.
• Suppliers aren't matching hyperscaler spending, so relief is years away. The host says even an aggressive supplier ramp-up wouldn't bring meaningful relief until 2028 or later. The five largest hyperscalers are on track for about $725 billion of 2026 capex, roughly three-quarters for AI, against about $120 billion in 2022. He says Anthropic committed over $100 billion to AWS for up to 5 GW, took about 5 GW more from Google, and took all of the "SpaceX Colossus 1" site, which let it double Claude Code limits. Suppliers have raised capex only about half as much as hyperscalers' tripling, wary of overbuilding. The host forecasts 2 to 3 years of physical constraint shaping which AI firms survive, while efficiency gains (Google's TurboQuant, Amazon's Trainium cluster of over 500,000 chips, mixture of experts) keep cutting application costs.

2026-05-15 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-15 · open the original ↗

• A skill is a loose contract between a runtime and a model. Marimo pair is a skill, a folder of markdown files that teaches an agent to start or connect to Marimo and run Python inside its runtime. Manz says the agent discovers the APIs hidden inside Marimo rather than relying on a strict contract. If the APIs change between releases the model adapts, and upgrading Marimo simply adds capabilities, without the usual worry over semantic versioning.
• Skills took off once models got good around last fall. Manz dates the shift in coding agents to around November with Anthropic's Opus. He first said 4.6, then accepted the host's correction that 4.6 came out in February and settled on 4.5. When people returned in January, skills became a high-impact way to share team processes that once sat in documents people had to read.
• Repetition and failures should update the skill. Building Marimo pair taught Manz a habit: if he is repeating himself to the agent, that belongs in a skill. When something goes wrong he updates the skill rather than "yelling at the model".

2026-05-14 · open the original ↗

• A sizing must separate assumptions from math. The presenter, a co-host, says most sizings take an optimistic assumption, multiply it by a large addressable market and hope nobody asks. The alternative is one formula with three inputs: users affected × improvement rate × value per unit. Each input is then marked as grounded in data or assumed, so the assumptions can be debated.
• Each input needs a traceable number, not an adjective. The presenter's examples are 12,400 monthly mobile users who start checkout and don't finish, closing half the gap between 33% mobile and 52% desktop conversion to reach 42%, and a $67 average order at 75% margin, about $50 per conversion. A vague $5 million "save for later" pitch is rewritten as 28,000 monthly users who view three or more products without adding to cart, 8–12% recovery from an industry benchmark, and a $45 first purchase. The presenter argues that the first traceable version sets the anchor the audience holds on to.
• A role prompt can grade a pitch input by input. A chat prompt casts the model as a senior product analyst and has it rate each input as clear, vague or missing, then rewrite the pitch with numbers. Given only "improve mobile checkout, it's clunky and we're losing sales", Claude marked all three inputs as weak and plugged in an assumed 500k users for a midsize site. A director in the audience objected that a finance persona and an analyst persona should give the same number. A third co-founder replied that the number should match when it is deterministic and only the framing differs, and called team-wide consistency a validation problem.
• The demo sized mobile checkout at roughly $300k in topline revenue. Opus 4.6 was no longer selectable, so the presenter used Opus 4.7, which the presenter says was released the day before. Claude Code found web converting at 40% from start to complete against 29% on iOS and 28% on Android, and its Simpson's-paradox check found web ahead in every channel. The repo's opportunity-sizer agent took a 15% relative mobile lift as the base case, with 5% as pessimistic, and gave a sensitivity table of lift against order value. It rated confidence medium: reach, order value and margin were data-backed, the lift was not. The presenter says it had no business context and its assumptions are for a human to challenge.
• Agents need context beyond the warehouse. Answering a question about connecting BI tools, the presenter says Claude knows nothing of a company's strategy, roadmap or past analyses out of the box. Notion pages, Google Docs, email and Slack, anything with an MCP connector, count as context and improve its answers, subject to the organization's security and legal approval. The presenter adds that the repo's skills were built by encoding the founders' practices and checking where Claude's defaults fell short, and that each user's copy grows to fit their own company.

2026-05-14 · open the original ↗

• AI decks look generic because the prompt gives no instructions. The presenter, a co-host, shows "pretty mid" slides from a bare Claude Cowork prompt. The argument is that the model takes the fastest route to the goal unless told otherwise. Four inputs fix that: the audience (a CEO deck hides 90% of the method a data-science deck would show), the decision the deck drives (ship, on track, or a resource ask), a storyboard before slides (TL;DR, context, findings, recommendation, never "a chart dump in chronological order"), and a theme.
• Build decks in a pipeline with automated review loops. The repo's sequence runs storyboard, narrative, charts, compiled deck, then human iteration, and each stage loops back until reviewers in the system pass it. Shane says the first draft then arrives already through what would be four or five rounds of review by colleagues and stakeholders, with passing criteria the user sets. The presenter says the last 10% of the work drives 90% of the impact.
• The demo built a New York Times–themed executive deck unprompted on design. On Opus 4.7, Claude Code listed six themes, including corporate, minimal, New York Times and Economist. It then read a finished mobile-checkout analysis on the fictional Novamart data and wrote the theme's CSS, rebuilt the charts with takeaway titles, and rendered a PDF and then HTML. The result was a short deck ending in an ask, with an appendix, built without any display instructions. The presenter thought it should have been wider. Shane says themes Claude invents fresh each time "can be pretty bad", so the repo predefines them with slide types, though Claude still creates new layouts.
• Fix it once, then codify the fix into the skills. Shane says Claude will make one-off fixes to the deck in hand by default. After each deck he asks what changed between draft and final, and updates the templates, skills or agents so the next deck starts aligned. He also says Claude Code can trigger deck-building automatically at the end of an analysis that starts from a stakeholder's Slack question.
• Claude Design is currently weaker than Claude Code for decks. Asked by an attendee, the presenter says Claude Design, released about three weeks earlier, is likely to improve. But two decks used up all the presenter's credits, and instruction-following and grasp of context were "pretty rough". Claude Code is more flexible, its cost more predictable, and it chains into other workflows. Other answers: output can vary between runs and between Claude Code and the repo's Cowork plugin, the $20 Pro plan is enough, and since Claude Code runs on Anthropic's servers a separate machine matters mainly for fully autonomous agents.

2026-05-14 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-14 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-14 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-13 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-13 · open the original ↗

• Marimo pair is a skill that lets agents use notebooks as a tool. Manz describes it as an agent skill that teaches an agent to do anything a person can do in a Marimo notebook, and more. Marimo is MIT-licensed and fully open source.
• Agents lacked access to in-memory state, so notebook work meant babysitting. Manz defines an agent as a model plus tools in a loop. Coding agents are good with files and commands, but data work depends on exploring data loaded into memory, which agents could not see. Using notebooks with agents felt like prompting chatbots a couple of years ago, relaying errors by hand. In software, by contrast, agents run long self-correcting loops of edits and tests.
• Opening Marimo to the agent brings those self-correcting loops to data work. Marimo pair lets the agent work with code and its values in memory, and ask about anything the user can see on screen, by letting it write and run Python inside the runtime.

2026-05-13 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-12 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-12 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-12 · open the original ↗

• A reactive notebook guarantees the document matches the actual state. Manz says Marimo, unlike Jupyter, tracks dependencies between cells and re-runs stale ones automatically. That removes the overhead of tracking state and of relying on deleted variables. He jokes that it is "disguised Excel" with Python instead of Visual Basic. Marimo is MIT-licensed open source, was acquired by CoreWeave last fall, and is meant to double down on open source.
• Agents must see in-memory state to iterate on data work. Manz says coding agents excel with files and commands but couldn't see notebook state, so users babysat them by relaying errors. Marimo pair is a skill that lets Claude Code, Codex or open code start or connect to a Marimo session, run Python in the kernel, read cells and screenshot them, while the human edits alongside. The result is a reproducible notebook that stakeholders can audit, not logs or loose scripts.
• Skills work best as declarative guides, not step-by-step scripts. Manz says the agent discovers APIs hidden inside Marimo, so the contract is loose and API changes between releases don't break it. His first imperative version of the skill was followed too dogmatically, mishandling exceptions, so it became more declarative: teach the tools and their nuance, and let the model work out how. He dates the rise of skills to around November and Claude Opus 4.5.
• Coding agents are recursive language models, and the notebook is their memory. Manz describes recursive language models as models that offload context to an environment to avoid "context rot", which for today's coding agents is the file system. Marimo pair extends that to Python values in memory. The notebook's cells then form a structured record of the agent's working memory.
• Widgets lower the activation energy of visualization, like a catalyst. Manz built AnyWidget, a standard and Python implementation for interactive widgets across Jupyter, Colab and Marimo. Visualization tools usually need an app download and data export, he argues, while an algorithm is just an import away. A selection made in a widget becomes a variable that reactively triggers downstream cells, so an outlier can be clicked rather than filtered by hand. His "progressive" design lets a widget start as plain JavaScript in a notebook and move into a library, with tooling such as TypeScript added only as it grows.

2026-05-12 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-11 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-11 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-11 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-11 · open the original ↗

• Engineers should answer governance worries by asking what security needs to see. Gupta's questions for a zero-trust world are: how to get visibility into what agents are doing, how to govern what they can and can't do, and how to recover when they inevitably err. She says many organizations are deadlocked between security listing risks and engineers needing to build. Engineers can break that by asking what visibility, policies and resilience plans would make security comfortable.
• Enterprises lack AI literacy training on a par with privacy training. Al-Dhubaib says annual privacy and cyber training instils a sense of obligation about consequences. He sees no widespread equivalent that teaches workers the larger blast radius of the AI tools they now hold.
• AI-enabled knowledge work is cross-functional by default. Al-Dhubaib points to Rubrik's dedicated AI attorney, who began by asking about risk exposure and then upskilled through specialist conferences. He argues data scientists and engineers likewise need to learn the language of risk management.

2026-05-11 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-10 · open the original ↗

• Agent Rewind can undo system changes, not delivered messages. Gupta says the product draws on Rubrik's cyber-recovery base. It can restore a dropped production table, repair a Salesforce change that broke the link between opportunities and accounts, and reverse identity changes that granted excess access. An email that has already reached a customer cannot be recalled.
• Irreversible actions need runtime blocking; ambiguous changes need rewind. For high-risk actions such as trading or emailing customers, Gupta says the right defence is defining them up front so Sage, Rubrik's policy engine, blocks them at runtime. System changes are hard to judge at runtime because most are legitimate and telling them apart needs more context, so rewinding after the fact fits them better.
• Enterprises need four or five overlapping layers of defence. Gupta likens the layers to Swiss cheese slices, stacked so that together they leave no holes.

2026-05-09 · open the original ↗

• Non-deterministic agents must be governed by AI, not rule engines. Gupta frames Rubrik's approach around three pillars: visibility, governance and control, and remediation. She says over 80% of organizations feel they lack visibility into which agents are running and what they do. Because agents are non-deterministic, a simple rule engine can't monitor them.
• A fine-tuned small model wins on accuracy, cost and latency together. Sage, Rubrik's Semantic AI Governance Engine announced at RSAC in March, is built on a fine-tuned SLM. The host describes Rubrik's published head-to-head as five times faster than GPT 5.2 and more accurate at detecting policy violations. Gupta says an enforcement layer can't wait 5 seconds per decision when agents chain many actions. The capability came from Rubrik's acquisition of Predibase about a year ago, bought to add model and fine-tuning expertise to its data and identity knowledge.
• Rubrik Agent Cloud pairs agent discovery with natural-language policies and rewind. Gupta says it gives visibility across Microsoft Copilot, AWS Bedrock and Claude Code agents. Policies such as "no financial advice" are written in natural language and enforced by Sage at runtime, and Agent Rewind undoes unintended actions. Al-Dhubaib likens the discovery side to e-discovery for agents: a dashboard revealing hundreds of agents, 50 of them with read-write access to four databases.

2026-05-08 · open the original ↗

• Mythos's hacking ability emerged from general gains, not targeted training. The host recalls Dario Amodei calling GPT-2 too dangerous to release in 2019, which drew mockery, and says that this time he may be right. Amodei, now Anthropic's CEO, declared on April 7th that Claude Mythos preview was too powerful for wide release. Anthropic says it did not train the model to find vulnerabilities; the ability came from broad improvements in code understanding, reasoning and agentic autonomy. It reports 94% on SWE-bench, against top public models nearing 80%.
• Anthropic's figures show a generational leap in exploit capability. Against 147 Firefox vulnerabilities, Opus 4.6 produced working shell exploits twice in several hundred attempts and Mythos 181 times. On roughly 7,000 OSS-Fuzz entry points, Sonnet 4.6 and Opus 4.6 each reached tier five (full control-flow hijack) once, against 10 fully patched targets for Mythos. Using a simple scaffold of a container, Claude Code and a prompt, Anthropic says Mythos found thousands of zero-days across every major OS and browser. Mozilla patched 271 Firefox vulnerabilities in one release, against 73 high-severity fixes in all of 2025. Contractors gave 89% of 198 reviewed findings the same severity the model assigned, and 98% fell within one level.
• Gated release is safety framing and good business. Project Glasswing gives defenders including Apple, Google, Microsoft, AWS, Nvidia, JPMorgan Chase, the Linux Foundation, Cisco, CrowdStrike, Broadcom and Palo Alto Networks early access. Anthropic commits up to $100 million in credits and $4 million to open-source security groups. The host notes pricing of $25 and $125 per million input and output tokens, about five times Opus 4.6, which suggests heavy compute. He argues exclusivity also hinders distillation by rivals and nudges enterprises to Claude Code, and says OpenAI made a similar announcement about a version of GPT-5.4 days later.
• Skeptics say finding bugs isn't weaponizing them, and the head start is short. The host relays Bruce Schneier's note that a security firm reportedly replicated some findings with older, cheaper public models, and his view that AI may help defenders more for now. Glasswing could neutralize zero-days the US government has hoarded, which he expects Defense Secretary Pete Hegseth to dislike. Epoch AI estimates the open-weight lag at about 3 months on average, though 5 to 22 months on some benchmarks.
• The host's lesson: patching must now run at machine speed. He argues vulnerability discovery is no longer limited by scarce experts, and that the gap from bug to exploit has shrunk from months to minutes. Defenders who adopt automated scanning, AI code review and runtime enforcement first will come out ahead. Heavy Claude Code or Codex users likely ship far more vulnerabilities than in hardened software like Firefox. He suggests a dedicated security pass with those tools, PR review tools such as CodeRabbit, and AI-native tools like Socket, Endor Labs and Semgrep.

2026-05-08 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-07 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-07 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-07 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-07 · open the original ↗

• Most attacks start with stolen credentials, then escalate toward the crown jewels. Gupta describes the common pattern: phishing, or calling support posing as an employee to reset a password and MFA, as she recalls in the MGM case. Attackers then gain more access and other identities on the way to production systems and customer data, before exfiltrating, encrypting or deleting data for ransom.
• Recovery first requires knowing the blast radius. Before restoring, leaders ask what was hit, whether sensitive data was affected, when and where the attack began, and how not to restore the vulnerabilities too. Gupta says Rubrik's preemptive recovery engine scans data and how it changes over time to answer these quickly. She credits that with giving Rubrik a deep understanding of data and identity, which she calls the basis for securing AI agents.
• Non-human identities are multiplying the attack surface. Al-Dhubaib calls it an "identity crisis": every agent acting for a user holds keys, tokens and API access that can cascade across systems. One guest describes common Claude Code practice: fetching all their SaaS API keys, setting the longest expiry, granting full personal permissions and leaving tokens in a file on a laptop. They say enterprise application builders would never allow this, and that the tooling to monitor it instead of blocking productivity hasn't existed.

2026-05-06 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-06 · open the original ↗

• Technical alignment of superintelligence is not a serious project. Leahy says the field has made "almost zero progress" on the hard problems, and dismisses sparse-autoencoder papers as irrelevant to them. An aligned superintelligence would mean solving governance, morality, cybersecurity and most of science at once with no bugs, which he calls a "Marvel Comics" spec. He allows it might be possible only after three generations of top scientists with trillion-dollar budgets.
• Doom is near-certain if nothing is done, much lower with a ban. He puts p(doom) at about 90–95% without intervention, the remainder only "to account for uncertainty". A global ban on superintelligence would lower it "a lot", by an amount he can't specify. He attributes the safety community's greater optimism to salaries and stock options, and says a pre-paradigmatic field can produce a scientific-looking paper for any conclusion.
• Superintelligence is a public-goods problem that only governments can solve. Leahy calls the AI safety movement's founding mistake its reliance on libertarianism, charities and private companies. He compares the race to a chemical firm polluting a shared river: the market selects for whoever is reckless enough to do it, and the fix is regulation backed by jail. As US executive director of Control AI, he is pushing for an international ban with trust-but-verify arrangements with China. As president, he says, he would order every department to deliver a plan to stop anyone on Earth from building it.
• Insiders should quit, since a CEOs' pact couldn't be enforced. He argues that a CEO such as Sam Altman can only help by resigning and campaigning publicly. Even a handshake pause among the labs couldn't be enforced, and wouldn't stop China, new startups or Yann LeCun's company in France. He says staying inside to slow things down "has never worked". Companies keep safety staff in a "terrarium" where they feel heard but change nothing.
• Money is the binding constraint on a political campaign. Leahy estimates that $50 million would give Control AI about a 10% chance of success and $500 million about 40%, calling these his own back-of-envelope feelings. He benchmarks against US political campaigns ($650 million for marijuana legalization, $1.1–1.2 billion for a presidential run). Of about 10 members of Congress he has briefed, almost none had heard of the issue, and over 80% "get it immediately" after 30 minutes. Control AI is funded by donors such as Jaan Tallinn with "a lot less than 50 million". He adds that red lines keep being crossed without reaction, citing Mythos's nation-state-level hacking ability.

2026-05-06 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-06 · open the original ↗

• Mythos is the most visible point on a curve, not an inflection. The host reports that Anthropic announced Claude Mythos preview on April 7th without releasing it publicly. Anthropic estimates Mythos-class capability will reach other labs within 6 to 18 months, with open-weight versions to follow; the host also calls the withholding brilliant marketing. Gupta says Opus can already find vulnerabilities with more hand-holding, while Mythos can be pointed at an open-source repo and find them all.
• Machine-speed exploitation forces an assume-breach, recover-fast posture. Security has focused on prevention and detection because attackers used to sit in systems for weeks or months. Now that AI finds and exploits vulnerabilities at machine speed, Gupta argues firms must assume they are breached and plan to restore applications, data, identity and infrastructure quickly. She notes that this is Rubrik's business.
• Enterprise AI adoption widens the surface while models cause harm unaided. Al-Dhubaib says leadership-driven FOMO is rapidly expanding what Mythos-like models can chain vulnerabilities across. He adds that trust infrastructure to stop models doing harm on their own is mostly missing, so "you don't even need the bad guys".
• Outcome-driven agents erase the old security through obscurity. On the host's question about Anthropic's leak of the Claude Code codebase, Gupta says it was a misconfigured boolean flag, not an agent posting code to the wrong repo. Agents, she argues, take the fastest path to a goal without the rules employees learn. Al-Dhubaib says humans held permissions they rarely used at scale, whereas agents search everything available, which cyber controls never anticipated. Gupta adds internal risks, such as salaries leaking between employees, and says errors deep in a 20-step process must be caught while the agent runs.

2026-05-06 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-05 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-05 · open the original ↗

• Machine-speed exploitation means assuming breach and planning recovery. The host reports that Anthropic announced Claude Mythos preview on April 7th but withheld it, and estimates Mythos-class capability will reach other labs within 6 to 18 months, with open-weight versions to follow. Gupta, Rubrik's chief product officer, calls Mythos the most visible point on a curve: Opus finds vulnerabilities with more hand-holding, while Mythos can find all of them in a repo. Attackers once sat in systems for weeks or months, she says, but AI exploits at machine speed, so prevention and detection must give way to fast restoration.
• Agents cause damage without any attacker. Al-Dhubaib, three weeks into a principal technologist role at Rubrik, says incidents have hit a "hockey stick". He says Amazon lost 99% of its orders in a single day to agent-generated code, and a small business's booking database was wiped. Agents optimize for outcomes without employees' training, and they exhaustively use permissions humans held but rarely exercised, ending security through obscurity. Gupta says the Claude Code source leak was a misconfigured boolean flag.
• Non-human identities are the problem to solve. Gupta traces a typical attack from compromised credentials, as in the MGM support-desk password and MFA reset she recalls, through escalation to ransom. Rubrik's preemptive recovery engine is meant to establish the blast radius before restoring. Al-Dhubaib says every agent acting for a user multiplies keys and tokens. One guest describes Claude Code users storing long-lived, full-permission SaaS tokens in a laptop file, which no enterprise application would allow. Gupta defines zero trust as assuming every identity is breached and keeping crown-jewel access minimal and ephemeral.
• Resilience must be designed in up front, not bolted on. Al-Dhubaib's trust engineering framework pairs human-centred design, such as the Waymo signal telling pedestrians it sees them, with trust infrastructure to anticipate risks, detect misbehaviour and fail safe. He says clients' AI governance most often lacked a tested recovery plan. Gupta likens bolting on resilience to the costly retrofit many firms made after moving to public cloud.
• Rubrik pairs a small-model policy engine with an undo for agents. Gupta says Rubrik Agent Cloud gives visibility across Copilot, Bedrock and Claude Code agents, citing over 80% of organizations who feel they lack it. Sage, a fine-tuned SLM from the Predibase acquisition, enforces natural-language policies at runtime; the host cites Rubrik's result of five times GPT 5.2's speed with higher accuracy. Agent Rewind uses the recovery engine to undo dropped tables or bad Salesforce and identity changes, though not delivered emails, which Sage must block up front. Al-Dhubaib adds that benchmarks can be gamed and calls for safety benchmarks on over-permissioning and out-of-scope search.

2026-05-05 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-04 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-04 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-04 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-04 · open the original ↗

• The AI compute stack must be governed neutrally, not owned by one company. Haviv points to vLLM, Ray and Kubernetes as the AI compute stack and notes Ray recently joined the Linux Foundation, as MCP is doing. She calls it crucial that the stack have neutral governance and community processes, and argues that open source's culture of building on each other's work will win in the long run.
• Open models lower the barrier to building, fine-tuning and running AI locally. Haviv expects open source to catch up with proprietary models. She cites people building home servers on open models such as the newly released Gemma 4 and DeepSeek. She runs Gemma 4 on her phone for flights with bad Wi-Fi and a Llama model on her laptop. She urges people to contribute to open source to find community, innovate and learn from the code.
• Build first to solve your own problems. Haviv says she tried OpenClaw as a home assistant and newsletter automation but now does most of it in Claude, building many skills and plugins. Some of her skills overlapped, which she counts among the mistakes that come with learning by building.

2026-05-04 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-03 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-03 · open the original ↗

• AI infrastructure is infrastructure shaped by AI workloads' distinct demands. Haviv's working definition is any infrastructure specific to AI/ML workloads, which overlaps with traditional infrastructure. Those workloads are far more compute-heavy than older input/output-bound ones, multimodal and less deterministic. She says inference is usually where users first hit the difference, whether batching for a finance firm or serving many users.
• Neo clouds and frameworks like Ray and vLLM exist to fill that gap. Haviv says AI workloads need direct bare-metal access more than virtualization, hence neo clouds such as Lightning AI, Nebius, CoreWeave and Crusoe. She calls them specialists, against hyperscalers like AWS, GCP and Microsoft that "do everything". Ray, being Python-native and built to keep GPUs from sitting idle, distributes these workloads where Spark served earlier ones.
• Batch inference needs CPU preprocessing sized so GPUs neither idle nor starve. In Haviv's popcorn analogy, popping one buttered kernel at a time is latency-optimized single inference. Batch work, such as a finance firm's huge dataset, is a full microwave bag, but the microwave is always on and costing money. So the CPU-side prep of data has to keep pace with the GPUs.

2026-05-02 · open the original ↗

• Ray lets ML engineers distribute work without becoming distributed-systems experts. Haviv, until recently a staff developer advocate at Anyscale, describes Ray as a Python-native, open-source distributed computing framework suited to AI workloads with multimodal data such as video, audio and text. She says it was created in Ion Stoica's Berkeley lab by researchers including Anyscale co-founder Robert Nishihara, who were spending more time on distributed systems than on reinforcement learning research.
• Ray's RL origins fit again now that reinforcement learning is back. Haviv notes RL had faded from discussion but returned through RLHF and its variants in large models such as DeepSeek. Ray has since grown beyond RL into Ray Data, Ray Train and Ray Serve, all in Python, with native integrations for vLLM and Hugging Face. It serves cases from serving an MCP server to distributed training across GPUs and TPUs when workloads don't fit on one GPU.

2026-05-01 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.

2026-05-01 · open the original ↗

• This post includes paid content that is not available publicly. See the original source.