Trump handed Anthropic its top models back, Microsoft laid off 4,800 people to fund its AI push, and Zuckerberg admitted his agents aren't that agentic yet. Who's actually shipping?
Hi,
Need a little inspo?
Check out these projects to get the creative juices flowing…
Edible Cups - Welcome to the future of sustainable snacking.
Autonomous AI sponsorship agent for podcasters.
It offers a Netflix-style subscription that protects travelers across Europe, the UK, Turkey, Canada, and more with an AI flight checker that instantly verifies disruption eligibility.
Sanare reads your whole health story - your history, your patterns, your life - and connects you to practitioners who can find what others have missed.
oHive helps luxury e-commerce businesses increase conversions and customer trust by enabling instant, drop-in video on business websites without an invitation. It's like Pinterest with interactive video.
The galaxy’s largest podcast directory
Let’s go.
This week from the AI for Founders podcast on Spotify :
The IPO That Buys Companies Instead of Firing Anyone — LINK
Todd Furniss took AI Squared public on Nasdaq to acquire real operating companies, keep their management teams, and layer in Transformational AI instead of running pilots or cutting jobs.
Built Is Not Built to Scale — LINK
Jason Katz of Kindling Solutions says most vibe-coded builds collapse without a governance layer, and joins Ryan for a check-in on sponsorships, scar tissue, and a possible Denver panel brawl.
Snipd is the world’s best podcast streaming app // try premium for free as a sub from this newsletter!
Investor Corner
Are you raising? Reply here, and we’ll feature your deck.
An autonomous AI job search agent for senior tech professionals. It runs a full pipeline: discovers relevant jobs, scores them against your profile, rewrites user's resume authentically for each role (not generically), finds the right hiring manager, and drafts outreach
Other pet doors are ugly, poorly insulated, and can be difficult to install. Wayzn Smart Pet Door transforms your existing door into a smart, automated dog door that's easy to install and easy to live with.
DiningOut unites restaurants, food lovers, industry suppliers, and brands through a self-reinforcing ecosystem of live events, targeted media, and proprietary technology.
Rare, Pristine, Protected drinking water.
The only platform that can authenticate human-made content. Using patented algorithms that analyze how content is made, not just the output.
Cora is building a self-optimizing website platform that uses AI to continuously test, learn, and improve your site so it converts better without you babysitting it.
AI-powered training simulations to help healthcare learners practice complex interpersonal and communication skills. We focus on realism, repeatability, and scalability—making soft-skills training more accessible and measurable.
AI + Human Stylists → LEARN MORE
Personalized menswear styling subscription
Less shopping, more living - dress for success with Taelor → Save $30 with RYAN30
Tip of the spear:
Zuckerberg admits Meta's AI agents are behind schedule — LINK
Mark Zuckerberg told staff at an internal town hall that Meta's AI agents haven't progressed as fast as expected, four months after a reorg that cut 8,000 jobs and moved 7,000 more onto AI teams. He still expects "more significant benefits" within three to six months, but the miss lands right after a $145 billion infrastructure bet.
Microsoft cuts 4,800 jobs as Xbox loses a fifth of its staff — LINK
Microsoft eliminated 2.1% of its workforce this week, hitting Xbox hardest as four studios spin off and management layers get slashed from 14 to as few as three. The cuts land the same week as a fresh $2.5 billion AI deployment unit, which tells you where the money's actually going.
OpenAI offers the US government a 5% stake to calm Washington — LINK
Sam Altman has proposed handing Washington a 5% equity stake worth roughly $42.6 billion, with Anthropic, Google, and Meta expected to cede similar slices under the same sovereign wealth fund concept. It's the clearest sign yet that frontier labs see government equity as the price of political peace ahead of their IPOs.
Agility Robotics goes public without promising you a robot at home — LINK
Agility will become the first pure-play humanoid robotics company on public markets via a SPAC merger, backed by more than $300 million in booked, multi-year robots-as-a-service revenue. CEO Peggy Johnson is deliberately skipping the consumer hype that's inflated rivals like Figure AI to a $39 billion valuation.
Anthropic launches Claude Sonnet 5 days after getting its top models back — LINK
Sonnet 5 is now the default model for Free and Pro users, pricing agentic work at $2/$10 per million tokens while scoring close to Opus 4.8 on coding and knowledge work. The launch lands right after Washington lifted its export restrictions on Anthropic's Mythos and Fable models, ending an 18-day suspension.
Microsoft launches a $2.5B forward-deployed AI unit — LINK
Microsoft Frontier Company will embed roughly 6,000 engineers directly inside client operations, following Amazon's $1 billion version of the same play two days earlier. Anthropic and OpenAI beat them both to it in May, so founders now have four flavors of "we'll build it inside your company for you."
Google and Amazon's own numbers show AI's real environmental bill — LINK
Google's carbon emissions are up 25% and Amazon's are up 16% year over year, with both hyperscalers' own sustainability reports pointing at AI-driven energy use as the culprit. Neither company says it outright, but the math is getting hard to offset away.
The running list of tech layoffs where employers actually blamed AI — LINK
TechCrunch's tracker now spans Meta, Cisco, Cloudflare, Coinbase, Intuit, and PayPal, with AI named directly in layoff memos even at companies posting record revenue. If you're benchmarking your own AI narrative against the market, this is the cheat sheet.
Most plant-based athletes are under fueled.
Momentous is the only brand that meets every standard expectation.
Without compromise. My favorite protein and creatine here at 20% disco for readers.
Framework:
Lean Team, Big Roadmap
Most founders think shrinking the team and shrinking the roadmap have to happen together. Gaurav Agrawal's bet at Collective Health says otherwise: keep roughly 210 engineers, hand them AI-augmented development across the board, and use the multiplier to collapse an 18-to-24-month roadmap into six. The team doesn't get smaller. The output does the growing.
- Audit your roadmap for AI-compressible work before you touch headcount. Go through every initiative on your 18-to-24-month roadmap and flag the ones that are bottlenecked by repetitive engineering work, not by judgment calls. Those are your compression candidates.
- Give every engineer an AI multiplier, not an AI replacement. Roll out AI-augmented development tools as a productivity layer for the people already on the team. The goal is more shipped per engineer, not fewer engineers per roadmap item.
- Re-scope your timeline publicly before you re-scope your team. Once the multiplier is in place, cut the roadmap timeline in front of the company, the way Agrawal moved from 18-24 months to six. A public timeline change forces the org to actually operate at the new speed instead of quietly absorbing the gains.
- Protect the headcount you have as your competitive moat. In a market where every competitor is cutting engineers to fund AI, a stable, AI-augmented team that ships faster is a structural advantage, not a cost problem to solve later.
- Measure the compression, not just the output. Track roadmap-months saved per initiative, not just velocity or story points. The number that sells this internally and externally is the ratio between what the roadmap used to cost in time and what it costs now.
Agentic partnership acquisition for the world’s top podcasts. —> gethowdi.com
Keep it moving
Etched exits stealth with $800M raised and a working AI chip — LINK
The inference-focused chip startup says it's signed over $1 billion in customer contracts and starts shipping this summer. Backers include Jane Street, Peter Thiel, and Geoffrey Hinton.
The DeepMind trio who built a poker AI are now running hedge fund money — LINK
Prague-based EquiLibre Technologies is now valued at $500 million after applying its poker-playing reinforcement learning to trading billions in daily volume across the S&P 500 and Nasdaq. Reinforcement learning's killer app might just be Wall Street.
Anthropic bets on workflow, not a new model, to win over scientists — LINK
Claude Science bundles 60-plus scientific databases and a project-manager AI into one workbench, running on existing Claude models rather than a new one. One Allen Institute researcher says it cut a two-year literature review down to weeks.
OpenClaw finally lands on Android and iOS — LINK
The open-source AI agent that captivated the internet earlier this year is now a mobile app on both platforms. Founders who've been running it from a terminal just got a much easier on-ramp.
Gemini Spark, Google's agentic assistant, comes to Mac — LINK
Google's plan-and-build agent expands beyond its original platform, joining the same agentic-assistant race as Claude and OpenAI's offerings.
The father of the internet is finally retiring — LINK
Vint Cerf, 83, is stepping down as Google's chief internet evangelist after 20 years. His parting warning: AI agents talking to each other in plain English is a recipe for a very expensive game of telephone.
Japan unveils its sovereign AI plan — LINK
Japan is the latest country staking out its own AI infrastructure ambitions, joining a growing list of nations unwilling to depend entirely on US or Chinese models.
Sony will end physical PlayStation discs in 2028 — LINK
Physical media's slow death picks up speed as Sony sets a hard date for going all-digital on new releases.
Google ships Nano Banana 2 Lite for faster, cheaper image and video generation — LINK
A leaner version of Google's image model trades some quality for speed and cost, aimed squarely at high-volume creative workflows.
Apple's Hide My Email feature has a bug exposing real addresses — LINK
A researcher claims the privacy feature has been leaking the real email addresses it's supposed to mask, undercutting the entire point of using it.
Ashton Kutcher leaves Sound Ventures to launch a new VC firm — LINK
Kutcher is teaming up with Morgan Beller, the architect of Meta's failed Libra project, to invest in "a post-AGI world," with a focus shifting from AI labs toward the infrastructure and energy underneath them.
Lime begins life as a public company — LINK
The scooter company finally goes public after years of will-it-won't-it speculation.
Lucid Motors' CFO is out as its new CEO continues a leadership shakeup — LINK
The exit comes as Lucid's Gravity SUV fails to gain traction and the company brings in a fresh slate of executive hires to try to turn things around.
The US government says it got hacked again — LINK
Another federal breach adds to a growing pattern of government systems getting compromised, a live backdrop to Washington's own AI security concerns.
Cloudflare stops charging AI per crawl and starts paying per answer — LINK
Cloudflare is scrapping its per-crawl fee for AI bots in favor of paying publishers only when their content actually shows up in an AI-generated answer, after finding some crawlers hit sites tens of thousands of times per referral sent back.
Meta quietly launches a vibe-coded gaming app called Pocket — LINK
A low-key launch signals Meta testing AI-generated game content as its own product category rather than just an internal tool.
Google publishes how it catches AI spam at scale — LINK
A new Google system terminated 50,000 coordinated spam clusters across 130,000 channels in six months by hunting networks of accounts rather than grading individual pages. Separately, an analysis of 220+ sites running scaled AI content found more than half lost 30%+ of their peak organic traffic once caught.
A spokeswoman is suing a company for using AI to make new ads featuring her — LINK
Another likeness-and-consent fight lands in court as brands keep testing how far AI-generated ad content can stretch without a new contract.
WTF is AI poisoning? — LINK
Digiday's explainer on the emerging tactic of competitors seeding bad-faith reviews and misinformation to manipulate how LLMs describe a rival brand. Product and service guides account for up to 28% of AI search citations, making them a prime target.
Amazon's Leo satellite internet is finally ready to launch — LINK
Amazon's answer to Starlink is finally approaching commercial service after years of delays.
Inbox Alchemy grows your list and publishes your weekly founder newsletter.
Email is the only distribution motion you actually own . Apply Now
What I’m thinking about:
Stop chasing superintelligence, customers just want reliability — Clip
Uncover the AI tools (here) you didn’t ask for, but now you’re weirdly into.
Thanks for being here.
-Ryan
p.s. I write about Revops, Product, and Founder-led marketing on Linkedin , X , and my blog .
