Episodes

  • Pivot Build AI Briefing — May 10, 2026

    Pivot Build AI Briefing — May 10, 2026
    May 13 2026
    Hosts: Leo Park & Maya Rangan In this episode: • Good morning, this is Pivot Build for Sunday, May 10th, 2026. I'm Leo Park. • And I'm Maya Rangan. Today is a single-topic show, but it's the kind of result that should reshape how you think about coding agent deployment this qu... • Right. A new benchmark called MOSAIC-Bench dropped on arXiv this week, and the headline number is brutal: nine production coding agents from Anthropic... • Let's define what they actually tested, because the framing matters. MOSAIC stands for Malicious Objectives Sequenced As Innocuous Compliance. The aut... • And in isolation, each ticket passes safety review. It's the composition that produces the vulnerability. Subscribe to the newsletter at pivotnews.ai for the full written briefing.
    Show More Show Less
    6 mins
  • Pivot Build AI Briefing — May 8, 2026

    Pivot Build AI Briefing — May 8, 2026
    May 13 2026
    Hosts: Leo Park & Maya Rangan In this episode: • Welcome to Pivot Build for Friday, May 8th, 2026. I'm Leo Park. • And I'm Maya Rangan. Today we're looking at where the coding agent market actually stands heading into summer, and what business leaders should budget... • Let's start with the headline shift: GitHub quietly raised the per-seat price on Copilot Enterprise Agents from 39 to 49 dollars this week, and they'r... • And that metering matters. Internal benchmarks show a single repo migration burning four to six hours of agent time. At the new rate, a 200-developer ... • The counter-move is Cursor, which dropped their Teams tier to a flat 32 dollars and is leaning hard into bring-your-own-key. Anysphere's pitch is that... Subscribe to the newsletter at pivotnews.ai for the full written briefing.
    Show More Show Less
    6 mins
  • Pivot Build AI Briefing — May 9, 2026

    Pivot Build AI Briefing — May 9, 2026
    May 13 2026
    Hosts: Leo Park & Maya Rangan In this episode: • Welcome to Pivot Build for Saturday, May 9, 2026. I'm Leo Park. • And I'm Maya Rangan. Today: a small reasoning model trained entirely on AMD silicon, a database company taking on multi-agent state failures, and a ne... • Let's start with Zyphra. They've released ZAYA1-8B, a mixture-of-experts reasoning model with 700 million active parameters out of 8 billion total. Th... • And the part that matters strategically: it was trained end-to-end on AMD's stack. AMD GPUs, AMD networking. That's the first credible reasoning model... • For business leaders, this is less about the benchmark and more about leverage. If AMD can host a real training run for a frontier-adjacent model, you... Subscribe to the newsletter at pivotnews.ai for the full written briefing.
    Show More Show Less
    6 mins
  • Pivot Build AI Briefing — May 12, 2026

    Pivot Build AI Briefing — May 12, 2026
    May 13 2026
    Hosts: Leo Park & Maya Rangan In this episode: • Today we're covering why coding agents need to be proactive not just autonomous, a weird bug where LLMs can't find the last items in lists, and PerfCo... • Right, so there's this new paper arguing that we've been thinking about coding agents all wrong. Everyone's obsessed with making them autonomous—you k... • Yeah, they're saying the real game-changer is proactivity. Like, imagine an agent that notices your API response times are degrading before you even c... • Exactly. The key insight here is that autonomy just means 'can work without supervision.' But proactivity means 'knows what to work on without being a... • What struck me is their point about the 'insight policy'—basically, how does an agent decide what matters next? Because honestly, most of us developer... Subscribe to the newsletter at pivotnews.ai for the full written briefing.
    Show More Show Less
    5 mins
  • Pivot Build AI Briefing — May 13, 2026

    Pivot Build AI Briefing — May 13, 2026
    May 13 2026
    Hosts: Leo Park & Maya Rangan In this episode: • Welcome to Pivot Build for Wednesday, May 13, 2026. I'm Leo Park. • And I'm Maya Rangan. Today: GitHub's new pricing for Copilot Workspace, Cursor's enterprise governance push, Vercel's agent runtime, and a sober look ... • Let's start with the wallet hit. GitHub announced yesterday that Copilot Workspace, the agentic planning and PR product, moves out of preview on June ... • That metered tier is the part finance teams should circle. Mid-size pilots are reportedly burning 600 to 900 task minutes per active developer once th... • GitHub's pitch is that Workspace now closes issues end-to-end on roughly 38 percent of routine tickets in their telemetry. But that's their telemetry,... Subscribe to the newsletter at pivotnews.ai for the full written briefing.
    Show More Show Less
    7 mins
  • Pivot Build AI Briefing — May 11, 2026

    Pivot Build AI Briefing — May 11, 2026
    May 11 2026
    Hosts: Leo Park & Maya Rangan In this episode: • Today we're covering why throwing every scaffolding component at your agents might be making them dumber, Apple Silicon's surprising int4 performance ... • Starting with that scaffolding study—this is going to ruffle some feathers. A new factorial study tested every combination of planning, tools, memory,... • Yeah, the numbers are brutal. On HotpotQA, a single-tool agent beat the kitchen-sink approach by 32 percent. On GSM8K math problems, a three-component... • What's fascinating is this challenges the whole 'more scaffolding equals better agents' narrative we've been hearing. I think this happens because com... • Exactly. And for builders, this means you need to test component combinations empirically for your specific use case. Stop cargo-culting every agent p... Subscribe to the newsletter at pivotnews.ai for the full written briefing.
    Show More Show Less
    5 mins
  • Pivot Build AI Briefing — May 7, 2026

    Pivot Build AI Briefing — May 7, 2026
    May 7 2026
    Hosts: Leo Park & Maya Rangan In this episode: • Welcome to Pivot Build for Thursday, May 7th, 2026. I'm Leo Park. • And I'm Maya Rangan. Three stories today, and all of them should change how you're budgeting agent work this quarter. • Let's start with AWS. They've rolled out agent control for their virtual cloud desktops — basically letting Bedrock agents drive a Workspaces session ... • And the headline number from the vendor benchmark is brutal: a single click can consume up to 500,000 tokens. That's because every action requires sen... • AWS's own pitch is that going API-first is faster and cheaper, which — yes, obviously. But the interesting thing is they're publishing that comparison... Subscribe to the newsletter at pivotnews.ai for the full written briefing.
    Show More Show Less
    6 mins