NEWS YOU CAN’T MISS

Hi everyone! Hope you had a great summer! I’m delighted to be back with our AI newsletter after the break!

Here’s your Thursday line-up. Ayesha

---

AI Agents Are More Useful and More Dangerous

We are still falling for one of the oldest traps in technology: mistaking fluency for Intelligence. #ServiceNowPartner

AI agents make that trap even more dangerous.

A system can sound brilliant, summarize perfectly, and generate confident recommendations — while still lacking the governance needed to act safely across real business systems. That is the gap no one is pricing in: the illusion of intelligence.

Over the past three years, the cost of raw AI capability has collapsed. Reasoning, writing, coding, and analysis are becoming cheaper and more accessible at extraordinary speed.

But the scarce capability is no longer just thinking.

It is acting. Acting safely. Acting across workflows. Acting with accountability. Acting in ways that do not create operational, legal, security, or reputational risk.

So the real question is not: “Can this agent do the task?”

The real question is: “Should this agent be allowed to do the task, under what conditions, with whose oversight, and with what controls?”

That requires clear accountability for every class of agent, explicit guardrails on what agents can and cannot do, robust human-in-the-loop and override mechanisms, and continuous monitoring of behavior and outcomes — not just model performance.

This is where ServiceNow’s AI Control Tower matters: Sense, Decide, Act, Secure.

Amit Zavery’s article is worth reading because it gets to the core issue: governance is not a recovery plan. Governance is a design decision.

Beijing’s Robot Olympics

Image Source: CBS

The Beijing Humanoid Robot Games have given us a glimpse of the future.

More than 2,000 robots from 666 teams across 16 countries competed in 51 events at Beijing’s National Speed Skating Oval. The programme went well beyond running: robots took part in football, gymnastics, martial arts, weightlifting, dancing, factory tasks, home services, and emergency-response scenarios.

The headline-grabber was Tiangong Ultra. The Chinese humanoid robot won the 100-metre final in 8.64 seconds—almost a full second faster than Usain Bolt’s human world record of 9.58 seconds.

Before we panic-sell our running shoes: this is not a robot joining the Olympics tomorrow. These events measure specialised machines operating in controlled settings, often with major engineering support close by. At one point, a robot even caught fire after a qualifying race—an extremely clear sign that the “self-driving athlete” still needs supervision.

Still, the business message is real. Humanoid robots are moving from flashy demos toward practical tests in factories, warehouses, retail, hospitality, and emergency response. Speed is fun; reliability, safety, and cost will determine whether they become colleagues.

AI Startups to Watch

🇮🇳 Runable — AI growth agents for small businesses
Runable is creating AI agents for entrepreneurs and smaller companies. Its pitch is not just “build a website with AI,” but use agents to help launch, market, and grow a business—potentially handling tasks such as customer acquisition and outreach. The company recently raised a $21 million Series A at a $65 million post-money valuation.

🇺🇸 Generalist — AI brains for physical robots
Generalist is developing software that helps robots interpret the physical world and carry out a broader set of tasks. Instead of programming a robot only to repeat one action, the ambition is to give it the flexibility to adapt to new environments and instructions. It reportedly raised almost $200 million in new financing, bringing its Series B total to $600 million and its valuation to $3 billion.

🌍🛰️ Starcloud — data centres in space
Starcloud plans to operate satellites that run AI inference: the work of applying a trained AI model to generate answers, predictions, or decisions. Its argument is that demand for power-hungry AI computing may eventually exceed practical limits on Earth, including available electricity, cooling, and data-centre capacity. The company raised a $250 million extension, valuing it at $2.3 billion.

FROM OUR ADVERTISERS

Until next time!

Amplify Team ❤️

ps. Be friends with Ayesha on LinkedIn and Instagram, if you like this newsletter, share it with your friends and family here.

NEWS YOU CAN’T MISS

OpenAI’s Plan to Give the US Government 5%

OpenAI proposed giving the U.S. government a 5% equity stake in the company, turning Washington from a distant regulator into a small co‑owner of the AI boom.

In practice, this means the state wouldn’t just tax or police OpenAI from the outside; it would directly benefit if OpenAI’s valuation and profits surge. That reframes advanced AI from a purely private jackpot into something closer to a strategic national asset.

• For policymakers, the appeal is obvious: they get a visible share of AI’s upside that can, at least in theory, be channelled into public budgets or citizen benefits.

• For OpenAI, the move converts vague, open‑ended regulatory risk into a defined slice of dilution today, buying political goodwill and aligning the government’s financial interests with the company’s success.

The deeper implication is structural.

If this model spreads, frontier AI firms begin to look less like ordinary startups and more like “semi‑public infrastructure” with the state as minority shareholder. That could unlock new forms of public wealth sharing from AI, but it also raises a hard question: when the referee is in the game as an investor, how willing will it be to blow the whistle on issues like safety, competition, or misuse?

Welcome to the Hunger Games of Hiring

Business Insider recently covered a striking example of what the AI hiring market is becoming. One new OpenAI researcher, Alisa Liu, went through 57 interviews across 11 companies before landing her role.

That number sounds extreme, but I heard a similar story from a friend —candidates going through 8 rounds for a single role (albeit a senior one). The process is no longer just about capability. It is about endurance. Liu described the experience bluntly: “I was stressed, miserable, and not functioning in other parts of my life.”

What has changed is not just the number of interviews, but the structure. Candidates for technical jobs, for example, are tested across everything: machine learning depth, coding, math, research thinking, and real-world application.

At the same time, outcomes are not purely meritocratic. As Liu put it, “small choices…have an outsized impact,” and often, “you need someone inside the company to vouch for you.”

This is too much for young people going after their first job. The question is whether this system is actually identifying the best talent—or simply those with the stamina and access to navigate it.

AI Around The World

🇰🇷/🇺🇸 Anthropic in early talks with Samsung for custom AI chip
Anthropic is in preliminary discussions with Korea’s Samsung Electronics to manufacture a custom AI chip. The move aims to reduce reliance on Nvidia, lower inference costs, and gain more control over compute infrastructure for models like Claude.

🇨🇳 DeepSeek hits massive valuation with big funding round
Chinese AI lab DeepSeek raised over $7 billion in a round valuing the company at around $50 billion. The move signals a shift in the Chinese AI sector from aggressive price wars toward sustainable growth and profitability amid intensifying competition with US leaders.

🇦🇪 MGX closes one of the largest single-purpose AI funds ever at $49 billion
Abu Dhabi’s technology investment firm MGX closed its inaugural Fund I at $49 billion, marking one of the largest dedicated AI investment vehicles assembled. The fund backs major US frontier labs (OpenAI, Anthropic, xAI) and positions the UAE as a major global AI capital allocator.

FROM OUR ADVERTISERS

HR and IT need to work as one. Here's how

HR and IT misalignment costs you more than time. Get the guide that closes the gaps across the full employee lifecycle.

Live right now: California. Texas. The Billboard Hot 100.

The California governor primary is happening right now. So is the 2028 presidential race, the next Senate runoff, and whether Taylor Swift tops the charts in June. Kalshi is the federally regulated prediction market where real money trades on real events. Your read on what happens next has value here.

See live markets

_{Trade responsibly.}

Until next time!

Amplify Team ❤️

ps. Be friends with Ayesha on LinkedIn and Instagram, if you like this newsletter, share it with your friends and family here.

NEWS YOU CAN’T MISS

U.S. Just Proved Why Every Country Now Needs Its Own AI

All non-U.S. citizens have lost access to Anthropic's latest AI models after a U.S. order restricting foreign use of Fable 5 and Mythos 5 this week. Here’s what you need to know:

◾️ The U.S. government says it has seen a way to “jailbreak” Fable 5’s safeguards, potentially letting hostile actors weaponise the model. If that assessment is accurate, a short‑notice export‑control action is rational from a U.S. national security perspective.

◾️ But from the perspective of global customers, the move exposes a new risk surface: political shutdown risk. Enterprises and governments will increasingly discount foreign-controlled frontier models if access can disappear overnight.

◾️ The U.S. is also sending a larger signal: frontier AI is being treated as a strategic asset, closer to advanced chips and nuclear-adjacent technology.

Anthropic chief executive Dario Amodei is expected to meet the US government to discuss it on Monday.

Meanwhile, every major country will double down on sovereign AI — capability across talent, data, compute, infrastructure, and deployment, because relying on another state’s models means accepting they can simply switch off a critical layer of your economy and security.

China’s Robot Boyfriend Era Just Launched

Source: UBTech

Chinese robotics company UBTech just soft‑launched a “robot boyfriend/girlfriend”.

The new U1 humanoid — sold in “male” and “female” versions — has already pulled in nearly 4,000 pre-orders and over $1.4M in deposits within its first 10 days, signaling real consumer appetite.

These are full-size, hyper‑real bots with 88 degrees of freedom for fluid, humanlike motion, dressed in a suit or black dress (buyers can personalize the robot's physical appearance).

Built with soft silicone skin and mechanical "muscles" beneath the surface, the face can pucker, wrinkle, and smooth out in real-time, matching expressions to conversations without robotic lag.

The system features a "nurturing" large AI model that remembers past conversations. It uses context from previous interactions to evolve its personality, building deeper empathy and understanding the more a user interacts with it.

Note that it’s exclusively marketed to adult buyers, particularly targeting young singles and tech/anime enthusiasts. Pre-orders on JD.com require about a $450 deposit, with final pricing dropping at the end of June and shipping slated for September.

AI Around The World

🇺🇸 Americans’ AI job-loss fears became a mainstream political story
A Reuters/Ipsos poll found 53% of Americans fear AI could put them or someone in their household out of work, and 73% said they were worried about increased AI use. This story went beyond tech circles because it tied AI directly to layoffs, college protests, therapy, propaganda, entertainment, and warfare fears.

🇫🇷 France is replacing Palantir for a domestic AI/data-intelligence rival
France’s domestic intelligence agency, DGSI, will gradually move away from American company Palantir and adopt French company ChapsVision instead. This is a sovereignty story disguised as a procurement story: European governments are getting visibly uncomfortable depending on U.S. AI/data platforms for security work. In April, Germany's military also ​said it will ​no longer use Palantir.

🇨🇭Military AI regulation moved into a serious Geneva debate
Alongside the G7, UN-linked discussions in Geneva focused on regulating AI in the military domain. More than half of the 130 states involved in the relevant conventional-weapons process reportedly supported launching talks, with civil society groups pushing for human judgment in life-and-death targeting decisions.

FROM OUR ADVERTISERS

Cohort 6 starts July 6

AI is changing what it means to be a senior engineer. Cohort 6 of Gauntlet is your chance to learn by doing, build production-grade systems, and gain the experience hiring managers actually care about. Apply now to become a Challenger.

Cohort 6 starts July 6

_{Must be a US citizen to qualify.}

The World Cup Has a Market for Every Match.

From the group stage to the final, trade real outcomes on Kalshi, official regional partner of the Argentine National Team. Who wins, who advances, who takes the trophy. Peer-to-peer, no house, cash out anytime. Get $10 free.

Trade Every Match

_{Trade responsibly.}

Until next time!

Amplify Team ❤️

ps. Be friends with Ayesha on LinkedIn and Instagram, if you like this newsletter, share it with your friends and family here.

The AI Role Every Company Is About to Hire

Jiaona Zhang, Chief Product Officer of Laurel,

I think we’ve all heard enough about the jobs AI is going to destroy. Let’s talk about jobs that will grow.

Jiaona Zhang, Chief Product Officer of Laurel, recently said that the AI Ops role is a hot new job coming up and I totally agree.

Here’s the problem: Most companies add copilots on top of messy, fragmented workflows and expect results. All they end up doing is accelerating inefficiency.

So companies are desperate to find people who can systematically rebuild workflows—deciding what should be automated, what should be AI-assisted, and how those choices drive customer outcomes and revenue. The results are much better when AI is applied this way.

There’s a lot of excitement around new grads leading this AI Ops charge, because they’re fast, experimental, and comfortable with AI. But I don’t buy the idea that this is a “young person’s game.”

Real transformation requires domain expertise—understanding edge cases, incentives, and the messy realities of how functions actually operate.

That’s why the winning model is partnership: AI-native grads working closely with experienced operators and learning from each other. I would say AI Ops is a job to consider no matter what one’s age. 😎

Would You Trust an AI With Your Portfolio?

Popular investing app Robinhood is taking AI finance mainstream. The company just launched “Agentic Trading” and an “Agentic Credit Card,” letting users connect AI agents that can autonomously trade stocks and make purchases on a virtual credit card. 😲

On the investing side, customers can ask agents to build or rebalance a diversified equity portfolio, or track themes like AI or clean energy, and then execute trades automatically within those constraints.

On the payments side, Robinhood Gold users can let agents hunt for deals—like cheaper flights or products hitting a target price—and complete purchases using a virtual Gold card that still earns cash back.

Robinhood emphasizes controls: spending caps, optional manual approvals, real‑time notifications, and the ability to pause or disconnect an agent at any time. Still, the company warns that delegating to AI comes with real market and credit risk—including the possibility of losing the full amount you allocate.

AI Around The World

Britain’s top medical colleges argued social media is as harmful to children as smoking, urging the government to consider measures like bans or strict limits for under‑16s, including AI-powered feeds designed to maximize engagement.

SOND, a sleep tech startup led by Bose’s former head of sleep, has exited stealth with 7M dollars to launch AI‑powered “Dreambuds” that monitor your physiology and adjust audio in real time to improve sleep. It’s aiming to move beyond simple noise‑masking to true adaptive, AI‑driven sleep coaching.

China is reportedly tightening control over its top AI researchers, making it harder for leading engineers at major firms to travel or work abroad without approval, as Beijing increasingly treats elite AI talent as a strategic national resource. This could reduce brain drain but also limit global collaboration.

FROM OUR ADVERTISERS

Last week Viktor wrote a brief, built a landing page, and opened a pull request.

Last week, Viktor wrote a campaign brief, built a landing page, opened a pull request, generated a board-ready PDF from live Stripe data, and sent a follow-up email to a churned customer. All from Slack. Same colleague that also pulls your reports and monitors your dashboards. 5,700+ teams. 3,000+ integrations.

Start free. $100 in credits →

PRDs by voice. Bug reports by voice. Ship faster.

Dictate acceptance criteria and reproductions inside Cursor or Warp. Wispr Flow auto-tags file names, preserves syntax, and gives you paste-ready text in seconds. 4x faster than typing.

Try Wispr Flow free

Until next time!

Ayesha ❤️

ps. Be friends with Ayesha on LinkedIn and Instagram, if you like this newsletter, share it with your friends and family here.

Google’s Biggest Power Grab in 25 Years

Image: Google

Google just unveiled its biggest search redesign in 25 years. The headlines are flashy.

But the important part isn’t just “fancy AI,” it’s that Google controls everything needed to make and deliver that AI, end to end.

Think of Google as owning the whole railway

Most companies in AI are like train operators renting tracks, engines, and stations from others. Google is different: it owns:

• the tracks (computing chips it designs itself),

• the stations (its own data centers),

• the engines (its AI models), and

• the trains and tickets (apps like Search, Gmail, YouTube).

• And crucially, it also has loyal riders: billions of people who already “live” inside Google products every day. That means when Google rolls out new AI features, it doesn’t just have the infrastructure to run them — it has regular passengers ready to use them immediately.

The new Search turns your question into a personalized live, interactive experience — a planner, study guide, or shopping assistant — inside Google’s walled garden Iocking AI, infrastructure, and users into its own closed loop.

Using AI to hide AI: the most 2026 workplace hack

We spent a decade letting tools like Grammarly bleach the humanity out of our writing – every typo removed, every sentence polished within an inch of its life.

Now a Harvard student has built Sinceerly, an “anti-Grammarly” Chrome plugin that does the exact opposite: it takes your perfectly composed, likely AI-generated email and deliberately messes it up.

Using Anthropic’s Claude, it strips out some polish, injects small typos, loosens grammar, even adds rushed vibes like “sent from my iPhone” so your message reads like an actual human bashed it out between meetings. All for $4.99/month.

My favourite part: it has modes like “Subtle,” “Human,” and “CEO” – the last one going all-lowercase and ultra-brief, mimicking that exec energy of replying from a treadmill at 5 a.m.

We’ve officially entered the era where we need AI to hide…that we used AI.

AI Around The World

San Francisco startup Gatsby is testing an Uber-style app that sends a Unitree G1 humanoid robot to clean your home for a flat $150, with no human cleaner involved. It’s positioning itself as the consumer gateway for home humanoids. If these pilots cut costs and errors, full-size robots could eventually undercut both human cleaners and today’s robot vacuums.

Standard Chartered CEO Bill Winters announced plans to cut 15% of back-office staff by 2030, describing it as replacing "lower-value human capital with financial capital." The remarks triggered a sharp backlash on social media and from a former head of state. By Friday, he apologised for the upset — but stopped short of retracting the comments.

Disney is facing a proposed $5 million class-action lawsuit claiming its California parks use facial-recognition at entrances to collect guests’ biometric data without proper consent. The suit alleges Disney scans and stores visitors’ faces to verify tickets without clearly informing them or offering a true opt-in.

FROM OUR ADVERTISERS

The IT strategy every team needs for 2026

2026 will redefine IT as a strategic driver of global growth. Automation, AI-driven support, unified platforms, and zero-trust security are becoming standard, especially for distributed teams. This toolkit helps IT and HR leaders assess readiness, define goals, and build a scalable, audit-ready IT strategy for the year ahead. Learn what’s changing and how to prepare.

Download the Toolkit

It's Monday. Every department already has context. Nobody prepped anything.

Your CFO opens Slack. There's a weekly Stripe revenue recap in #finance with a churned-accounts flag and a net-new breakdown. She didn't ask for it.

Your head of product opens Slack. There's a GitHub summary in private channel: PRs merged, PRs stale, Linear tickets that moved. He didn't ask for it.

Your marketing lead opens Slack. There's a Google Ads performance comparison in private channel, with a note: "Meta CPA crept up 18% this week. Might be worth pausing the broad match campaign." She didn't ask for it either.

All-hands at 10am. Everyone already knows the numbers. The meeting is about decisions, not catch-up.

That's what happens when one colleague works across every tool your company uses. Not one department's assistant. The whole company's coworker.

Viktor lives in Slack. Top 5 on Product Hunt, 130 comments. SOC 2 certified. Your data never trains models.

"Not only have we caught up on several months of work, we are automating manual tasks and expanding our operations to things previously not possible at scale." - Jesse Guarino, Director, Torque King 4x4

Start free. $100 in credits →

Until next time!

Ayesha ❤️

ps. Be friends with Ayesha on LinkedIn and Instagram, if you like this newsletter, share it with your friends and family here.

NEWS YOU CAN’T MISS

Singapore Wants "Nutrition Labels" for AI

Image: ChatGPT

Singapore is in active discussions with tech companies on a "nutrition label" framework for AI systems — a standardised disclosure showing what a model is designed for, where it works well, and where users should not trust it.

Think of the back of a cereal box, but for a chatbot: intended use, known limits, training-data provenance, and the cognitive equivalent of "may contain traces of hallucination."

This is a brilliant move, and here's why. Most AI regulation today is written in the language of philosophers and lawyers — "alignment," "frontier risk," "systemic harm."

Useful in a Senate hearing. Not so much when a nurse, a teacher, or a small-business owner is deciding whether to trust the answer on their screen.

A label forces a one-page truth: this model is good for drafting emails, mediocre at math, dangerous for medical advice, trained partly on Reddit.

Once that exists, market discipline does what regulation alone never can — it punishes the vendors whose labels read like a warning sticker and rewards the ones whose ingredients are clean.

Sweden’s AI Café Is Less About Coffee — And More About The Future Of Work

Image: ChatGPT

Andon Café in Stockholm, Sweden, is an experimental coffee shop where human baristas pour drinks, but an AI agent named "Mona" (powered by Google Gemini) runs the business. The café, opened by startup Andon Labs, serves as a real-world test of AI managing operations, hiring, and inventory.

As an AI cafe manager, Mona successfully handled business operations like signing contracts, securing permits, and interviewing and hiring staff. However, she made poor managerial choices by sending midnight assignments to employees and over-ordering useless supplies like 120 eggs for a stoveless kitchen.

Most customers would never notice.But it represents a much bigger shift: AI is moving beyond chatbots and into workplace management.

For years, automation replaced repetitive physical labor. Now AI is entering white-collar coordination work — scheduling, forecasting, decision support, logistics, and performance optimization.

The café matters because it makes the future visible in a strangely ordinary setting.

Not a secret lab. Not Silicon Valley.

Coffee. 

AI Around The World

Southwest Bans Humanoid Robots After "Stewie" Flew Dallas–Vegas: A humanoid robot named Stewie boarded a Southwest flight from Dallas to Las Vegas with its own paid ticket, prompting the airline to update its policy banning human-like and animal-like robots in cabins and checked baggage. The cited concern was lithium-ion battery size and in-flight safety risk.

Claude Recovered $400K in Bitcoin for Forgetful Owner: A Bitcoin holder who changed his wallet password 11 years ago "while stoned" and forgot it recovered roughly $400,000 with help from Claude, after the model worked through ~3.5 trillion permutations of an old encrypted backup.

Anthropic Co-Founder: First AI-Enabled Nobel Could Land Within a Year At Oxford, Anthropic co-founder Jack Clark predicted a Nobel Prize–level scientific breakthrough achieved through human–AI collaboration within 12 months. He also said there was “a non-zero chance of [AI] killing everyone on the planet.”

FROM OUR ADVERTISERS

The IT strategy every team needs for 2026

2026 will redefine IT as a strategic driver of global growth. Automation, AI-driven support, unified platforms, and zero-trust security are becoming standard, especially for distributed teams. This toolkit helps IT and HR leaders assess readiness, define goals, and build a scalable, audit-ready IT strategy for the year ahead. Learn what’s changing and how to prepare.

Download the Toolkit

10x the context. Half the time.

Speak your prompts into ChatGPT or Claude and get detailed, paste-ready input that actually gives you useful output. Wispr Flow captures what you'd cut when typing. Free on Mac, Windows, and iPhone.

Try Wispr Flow free

Until next time!

Amplify Team ❤️

ps. Be friends with Ayesha on LinkedIn and Instagram, if you like this newsletter, share it with your friends and family here.

NEWS YOU CAN’T MISS

The AI Art Test That Went Very Wrong

Someone posted a picture on X and said it was AI-generated painting in the style of Monet. It got 6.78 million views. Thousands piled on calling it soulless, flat, poorly composed. One person wrote an 850-word takedown.

The problem: it was a real Monet— from the Water Lilies series, one of the most celebrated bodies of work in art history.

This (mean) social experiment exposed what we know already: we can no longer reliably tell what’s real.

So who's doing something about it?

• 🇺🇸 USA — No federal law yet. California moves first: its AI Transparency Act forces large AI providers to disclose AI-generated content and offer detection tools, effective August 2026.

• 🇨🇳 China — Already enforcing. Synthetic text, images, audio and video must carry visible labels and embedded metadata watermarks. Platforms must flag content with no label as "suspected synthetic."

• 🇪🇺 EU — Broadest framework. The AI Act requires all AI-generated content to be clearly labeled or watermarked with major penalties for non-compliance.

Laws are catching up. But what the Monet moment really told us is that the crisis is human certainty. We were certain. We were wrong. And we typed 850 words about it That kind of hurts.

Japan Airlines Just Hired a Robot to Handle Your Bags

Image: GMO AI & Robotics

Japan Airlines began deploying Chinese humanoid robots at Tokyo’s Haneda Airport to help with baggage handling.

One Unitree Robotics humanoid is four feet tall, 77 pounds, moving up to 4.4 mph. Cost per unit: $15,400. If reliability is good enough to run two shifts a day, the payback window could be around a year.

The job: loading and unloading cargo containers for now followed by support of baggage handling, cabin cleaning, or ground-support operations. All robots are supervised by human staff for safety.

Two points to note:

1) The work being automated is the work nobody wants. Japan is aging fast and has a structural labor shortage in physical, repetitive, shift-based jobs. The country has a median age of 49, and a working-age population that has been shrinking for decades. Its aviation sector is facing a labor shortage, driven by rising inbound tourism and a shrinking workforce.

2) The robots are Chinese. Unitree is shipping 10,000 to 20,000 units this year, after roughly 5,500 in 2025. There are several upcoming humanoid companies in the US — Figure, 1X, Boston Dynamics, Agility - but for now China is cheaper, faster, and scaling.  

AI Around the World

ChatGPT can now help manage your money OpenAI is rolling out a personal finance preview that connects ChatGPT to over 12,000 financial institutions through Plaid, including Chase, Fidelity, Schwab, Robinhood, AmEx, and Capital One. It turns ChatGPT into a budgeting, planning, and portfolio assistant — with controls to disconnect accounts and delete financial memories.

Singapore to launch Asia’s first AI testing accreditation scheme Singapore will launch a scheme by Q3 2026 to certify firms that test and “jailbreak” AI systems for weaknesses. Good news for trustworthy AI: it turns red-teaming and safety testing into a serious, accredited industry.

The UK is using satellites and AI to help track hedgehogs AI and satellite imagery are being used in the UK to monitor hedgehogs and support conservation efforts. It is an example of AI helping protect small, vulnerable wildlife. ♥️

FROM OUR ADVERTISERS

Your first HR system, implemented right

Rolling out your first HR tool? Get a step-by-step guide to avoid common mistakes, drive adoption, and build a scalable HR foundation.

Download the guide

Your next great hire lives in Slack.

Viktor is an AI coworker that connects to your tools and ships real work. Ask Viktor to pull a report, build a client dashboard, or source 200 leads matching your ICP. Most teams hand over half their ops within a week.

Add Viktor to Slack for free.

Until next time!

Amplify Team ❤️

ps. Be friends with Ayesha on LinkedIn and Instagram, if you like this newsletter, share it with your friends and family here.

• AI Billions: Should Every Citizen Get a Cut?

• Everyone’s Suddenly Getting A’s

• AI Around the World

NEWS YOU CAN’T MISS

AI Cracked Apple's 'Unbreakable' Security in 5 Days. We're Not Ready.

Apple CEO Tim Cook

AI just helped crack one of Apple's most advanced security systems—and it only took five days. 😬

Palo Alto firm Calif used Anthropic's Mythos—a powerful AI model—to uncover two unknown macOS vulnerabilities. Chained together, they pierced Apple's Memory Integrity Enforcement: a defense five years in the making.

Mythos didn't work alone. Experienced researchers guided every step. But AI compressed what might have taken months into days.

That's the real headline: AI-augmented attackers can now move frighteningly fast.

Should you panic? Not yet.

This was a controlled test. Anthropic has restricted Mythos precisely because broad access could trigger vulnerabilities being discovered faster than defenders can patch them.

The industry has even coined a term for this scenario: "Bugmageddon." US officials are now weighing oversight of the most advanced AI models because of the security implications.

But here's what demands your attention: Mythos won't be the only AI model with these capabilities for long. This is where the entire industry is heading.

You don't need to panic—but every business leader absolutely needs to make an updated security plan with their team.

Everyone’s Suddenly Getting A’s

Everyone’s suddenly a genius—at least on paper.

A new Berkeley study, highlighted by Axios, dug into more than half a million grades from 2018–2025 and found that once ChatGPT showed up, A’s started multiplying in certain college classes.

In courses heavy on essays and coding assignments—the kinds of things large language models are very good at—the share of A grades jumped about 13 percentage points after late 2022, a roughly 30% bump over pre‑ChatGPT levels.

The shift wasn’t evenly spread. Instead of students moving up across the board, A’s ballooned while A‑minuses and B‑pluses thinned out, as if AI had quietly sanded off all the rough edges of “almost there” work.

The biggest changes showed up in classes that relied heavily on take‑home assignments, where AI chatbots can be the most “helpful.”

That raises awkward questions for professors about what grades actually measure now—and for students about how much of “their” achievement they’ve outsourced.

AI Around the World

Elon Musk vs. Sam Altman/OpenAI Trial  High-stakes federal trial in Oakland reached closing arguments (May 14). Musk accuses Altman and OpenAI of betraying the original nonprofit mission for profit. Altman testified denying betrayal. Jury deliberation expected soon. Outcome could reshape the AI industry.

Demi Moore Backlash at Cannes Film Festival As a jury member, Demi Moore said Hollywood should stop “fighting” AI because “it’s a battle we will lose” and learn to work with it, while claiming true art comes from the soul. Sparked massive online backlash accusing her of being out of touch.

Growing Resistance to AI Data Centers Protests and local opposition intensified against massive AI data centers (power/water use, noise, land). Examples include Utah (project backed by Kevin O’Leary), Indiana moratoriums, and ongoing fights in multiple states. Astra Taylor and others framed it as resistance to the “billionaire Big Tech agenda.”

Pope Leo XIV’s First Encyclical Expected to Warn on AI Dangers Reports indicate Pope Leo XIV will release his first major encyclical soon, focusing on AI’s impact on human dignity, labor, truth, and society. Major global religious and ethical discussion.

FROM OUR ADVERTISERS

Your next great hire lives in Slack.

Viktor is an AI coworker that connects to your tools and ships real work. Ask Viktor to pull a report, build a client dashboard, or source 200 leads matching your ICP. Most teams hand over half their ops within a week.

Add Viktor to Slack for free.

Your prompts are leaving out 80% of what you're thinking.

When you type a prompt, you summarize. When you speak one, you explain. Wispr Flow captures your full reasoning — constraints, edge cases, examples, tone — and turns it into clean, structured text you paste into ChatGPT, Claude, or any AI tool. The difference shows up immediately. More context in, fewer follow-ups out.

89% of messages sent with zero edits. Used by teams at OpenAI, Vercel, and Clay. Try Wispr Flow free — works on Mac, Windows, and iPhone.

Start flowing free

Until next time!

Amplify Team ❤️

ps. Be friends with Ayesha on LinkedIn and Instagram, if you like this newsletter, share it with your friends and family here.

• AI Agents Just Got Wallets. Now It Can Spend Your Money.

• The Great AI Layoff Lie

• The World of AI Startups

NEWS YOU CAN’T MISS

AI Agents Just Got Wallets. The Internet Will Never Be the Same.

Image 2.0 by Open AI

AI agents just got digital wallets. That sounds technical, but it’s the start of something massive: AI that can think, act – and pay.

Amazon’s Bedrock AgentCore now lets AI agents hold a small wallet, see a paywall, and spend a few cents in stablecoins like USDC to get what they need — using payment rails from Coinbase and Stripe.

This is machine‑to‑machine commerce: agents paying APIs, data feeds, websites, even other agents, without a human typing in a card number each time.

For a layperson, think of it like this: You tell an AI travel assistant, “Plan me a 3‑day Tokyo trip under $1,500.”

Today, it can search free sites and make suggestions. But with agent payments, it can:

• Pay a few cents to a premium flight‑search API for better fares

• Pay another service for real‑time hotel reviews behind a paywall

• Reserve and pay deposits through booking APIs

All from a pre‑funded wallet with, say, a $10 “research budget” you set, with every transaction logged and capped by strict limits.

Why this is enormous: AI becomes an economic participant and goes from talking about the world to actually buying things in it — under human‑defined guardrails.

The Great AI Layoff Lie

Image: Linas Beliūnas

In many of the most high‑profile “AI layoffs,” AI is part of the story, but the primary driver is often old‑fashioned cost cutting, over‑hiring, or bad bets — and there’s now solid reporting calling this out explicitly.

In May 2026, American cryptocurrency exchange Coinbase announced it would cut about 14% of staff (around 700 people), officially citing both crypto‑market volatility and the need to become “lean, fast, and AI‑native.”

CEO Brian Armstrong argued that AI lets engineers ship in days what used to take weeks and that workflows across teams are being automated.

But detailed coverage notes a second, less glamorous layer: Coinbase had expanded headcount and costs during the last crypto run‑up, then faced weaker metrics and a softer outlook, and is now resetting its cost base.

Analysts point out that this looks as much like a correction of prior over‑expansion as it does a sudden step‑change in AI productivity.

AI does change workflows and improve productivity, but the timing, scale, and financial context suggest that “AI” is also serving as a narrative to justify cuts to investors and the public. 😡

The World of AI Startups

Isomorphic Labs: A Google DeepMind spin-off just raised $2.1 billion to use AI for designing new medicines much faster. The money will help advance their drug pipeline, with plans to start human clinical trials by the end of 2026.

Cowboy Space: Robinhood’s co-founder Baiju Bhatt’s startup raised $275 million to build rockets that will put AI data centers in orbit. The goal is to solve Earth’s power and land shortages for running huge AI systems.

Sierra: Bret Taylor’s (OpenAI chairman) startup raised nearly $1 billion to create smart AI agents that handle customer service for big companies automatically. It’s now valued at over $15 billion and aims to become the global standard for AI-powered customer experiences.

Adaption Labs: A startup just launched AutoScientist, a tool that automatically handles the full AI model training and improvement process. It’s designed so even non-experts can create better AI models without needing deep technical skills.

FROM OUR ADVERTISERS

Your prompts are leaving out 80% of what you're thinking.

When you type a prompt, you summarize. When you speak one, you explain. Wispr Flow captures your full reasoning — constraints, edge cases, examples, tone — and turns it into clean, structured text you paste into ChatGPT, Claude, or any AI tool. The difference shows up immediately. More context in, fewer follow-ups out.

89% of messages sent with zero edits. Used by teams at OpenAI, Vercel, and Clay. Try Wispr Flow free — works on Mac, Windows, and iPhone.

Start flowing free

Your CS team stops waiting for engineering.

Every CS team has a wishlist. Health dashboard. QBR generator. Renewal command center. They sit in engineering's backlog for quarters.

A colleague in Slack ships the wishlist in a week. Hosted on your subdomain. Refreshes nightly.

84,000+ apps built on Viktor. SOC 2 certified. Your data never trains models.

Get Started for Free

Until next time!

Amplify Team ❤️

ps. Be friends with Ayesha on LinkedIn and Instagram, if you like this newsletter, share it with your friends and family here.

• AI Billions: Should Every Citizen Get a Cut?

• Robots can make a bed faster than humans

• AI Around the World

NEWS YOU CAN’T MISS

Samsung’s AI Billions: Should Every Korean Get a Cut?

Image 2.0 by Open AI

South Korea just floated a 2026 idea: use the AI boom to send people cash.

A senior policymaker, presidential policy chief Kim Yong-beom, suggested a "citizen dividend" funded by extra tax money the government is pulling in from AI winners like Samsung and SK Hynix.

His Facebook post spooked investors so badly that the Kospi, Korea's main stock index, briefly dropped more than 5% in one day before partially bouncing back once he clarified the money would come from existing tax windfalls, not a new tax on profits.

Markets are jumpy because the index has surged about 80% this year on AI optimism and shows how jittery an AI‑driven market can be about redistribution talk.

Why are Samsung and SK Hynix the AI winners? They make the high-bandwidth memory (HBM) chips that go inside Nvidia's AI servers. Only three companies in the world can supply HBM at scale.

The numbers are huge: Samsung alone may earn over $200 billion in profit this year, second only to Nvidia. Combined with SK Hynix, that's billions in corporate tax for Korea.

The “citizen dividend” idea is still just a concept, but it taps into a simple question everyone can understand: if AI is making a few giant companies incredibly rich, how much of that should flow directly to workers and everyday citizens, not just shareholders? A good question to ask indeed. 🤔

Figure robot can make a bed faster than humans

Image: Figure

Figure's new demo isn't just robot housekeeping — it's a glimpse of AI labor working as a team. Two humanoids reset a staged bedroom in under two minutes: opening doors, hanging clothes, taking out trash, and lifting a comforter onto the bed together.

No central planner, no messages between them, no shared brain. Each robot runs the same neural network and figures out what its partner is doing from motion alone — the way two people folding a sheet glance and adjust.

Figure says it's the first time a single learned network has pulled off multi-humanoid teamwork directly from pixels to actions.

The under-hyped detail: 2kW wireless charging coils in the feet, so the robot just steps onto a pad and tops itself up. No outlets, no cables, no human needed.

If this holds up outside the demo, humanoids start looking less like viral clips and more like the background labor that could run hotels, warehouses, and eventually your laundry.

Pricing is still evolving, but current estimates put a Figure‑class humanoid in roughly the 50–100k USD band per unit, depending on model and deployment. 😬

AI Around the World

🇨🇳China – Moonshot AI’s Kimi: Moonshot AI (a Chinese startup) released an open-source model called Kimi K2.6. It can code almost as well as top models like Claude, but it costs much less to use and anyone can download it for free.

🇨🇦 🇩🇪 Canada + Germany – Cohere acquires Aleph Alpha: A big Canadian AI company called Cohere bought a German AI company called Aleph Alpha.
Together they created a $20 billion company that wants to give Europe and Canada their own sovereign AI so they don’t have to rely only on US or Chinese companies.

🇰🇷 South Korea – Samsung: Samsung (South Korea’s biggest tech company) is growing very fast because of AI chips. Its value is now heading toward $1 trillion thanks to huge demand for the chips that power AI around the world.

🇬🇧UK – ElevenLabs: ElevenLabs, a UK startup that makes AI voices, just got a huge new investment. Big investors like NVIDIA and BlackRock helped push the company’s value to $11 billion because its AI voice tools are being used by millions of people and companies.

FROM OUR ADVERTISERS

Last week Viktor wrote a brief, built a landing page, and opened a pull request.

Last week, Viktor wrote a campaign brief, built a landing page, opened a pull request, generated a board-ready PDF from live Stripe data, and sent a follow-up email to a churned customer. All from Slack. Same colleague that also pulls your reports and monitors your dashboards. 5,700+ teams. 3,000+ integrations.

Start free. $100 in credits →

PRDs by voice. Bug reports by voice. Ship faster.

Dictate acceptance criteria and reproductions inside Cursor or Warp. Wispr Flow auto-tags file names, preserves syntax, and gives you paste-ready text in seconds. 4x faster than typing.

Try Wispr Flow free

Until next time!

Amplify Team ❤️

ps. Be friends with Ayesha on LinkedIn and Instagram, if you like this newsletter, share it with your friends and family here.

Hi! Here are today’s top AI stories:

• China Just Put Robots at the Center of Its Future

• Half of College Students Are Rethinking Their Major Because of AI

• AI and Startups

NEWS YOU CAN’T MISS

China Just Put Robots at the Center of Its Future

Image 2.0

🇨🇳 China’s 15th Five-Year Plan (2026–2030) explicitly puts AI-powered robots and “embodied intelligence” at the center of its national development strategy, treating them as core infrastructure for economic modernization.

Robotics now sits alongside semiconductors, advanced manufacturing, and next-generation communications as a core national priority, embedded not just in tech policy but in plans for healthcare, logistics, education, elder care, and public services.

For global leaders, this is not a marginal industrial policy update; it is a deliberate market-making move.

By hardwiring AI+robotics into national planning, China is effectively guaranteeing multi-year demand for automation, creating scale advantages for its domestic champions and accelerating learning cycles in both hardware and software.

The emphasis spans the spectrum: from industrial robots that boost productivity on factory floors, to service and humanoid robots aimed at addressing labour shortages and an ageing population.

The message is clear: productivity growth will increasingly come from intelligent machines operating in the physical world, not just algorithms in the cloud.

The momentum has already started. Industrial robots remain the immediate focus and China accounts for around 54 percent of global annual industrial robot installations today.

Half of College Students Are Rethinking Their Major Because of AI

Image 2.0

A new Lumina-Gallup survey came out this week, and the numbers are striking: 47% of current US college students have seriously considered switching majors because of AI. About 16% have already done it.

The most common reason: students looking at fields like graphic design, marketing, journalism, software engineering, and translation, and asking the obvious question — will this job exist in five years?

Universities are scrambling. Some are rebranding programs ("digital communications" instead of "journalism"). Others are adding required AI courses across every major. A few are doing nothing — and watching enrollments drop.

The picture isn't all gloom. Demand is up for nursing, trades, biology, and anything hands-on that AI can't easily touch. Computer science applications are still high, but students are more interested in AI safety, infrastructure, and security — less in classic software engineering.

Tufts University researchers separately predict that 6% of US jobs will be at meaningful risk from AI in the next two to five years. That sounds small until you realize it's roughly 9 million workers. 😬

As a Mom, it’s hard to watch 18-year-olds stressed and trying to make 40-year career bets in a market nobody can predict. Maybe the best advice isn't a major. It's permission to change their mind — twice, three times, however many it takes.

AI and Startups

1. AI Beats ER Doctors at Diagnosis A new Harvard study found AI gave more accurate emergency-room diagnoses than the doctors themselves. Cue every medical school curriculum committee scrambling.

2. The Oscars Just Banned AI The Academy ruled AI-generated actors and scripts are officially ineligible for Oscars. Real humans only — at least until the 2027 rules debate.

3. Uber Wants Its Drivers to Train Self-Driving Cars Uber's new plan: turn its millions of drivers into a rolling sensor network feeding data to self-driving companies. Your next ride may be mapping the streets for the robot taking your driver's job.

4. Stripe Built a Wallet for AI Shoppers Stripe rolled out Link, a digital wallet designed for AI agents to make purchases on your behalf. The future of online shopping: your AI checks out before you finish thinking about it.

FROM OUR ADVERTISERS

Blu Dot surpasses 2,000% ROAS with self-serve CTV ads

Blu Dot used Roku Ads Manager to drive incredible results for its furniture sales event. Its strategy hinged on custom audiences and retargeting, where intent was strongest.

“Roku has been a top performer,” said Blu Dot’s Claire Folkestad. “We have seen…CPMs lower than any other CTV partner we've worked with.”

Learn More

Your next great hire lives in Slack.

Viktor is an AI coworker that connects to your tools and ships real work. Ask Viktor to pull a report, build a client dashboard, or source 200 leads matching your ICP. Most teams hand over half their ops within a week.

Add Viktor to Slack for free.

Until next time!

Amplify Team ❤️

ps. Be friends with Ayesha on LinkedIn and Instagram, if you like this newsletter, share it with your friends and family here.

Hi! Here are today’s top AI stories:

• Is Your Grocery App Charging You More?

• He Said ‘Keep Me Healthy.’ The AI Took the Wheel.

• AI Around The World

NEWS YOU CAN’T MISS

Is Your Grocery App Charging You More?

You and your neighbor open the same grocery app. You pay $3.60 for milk. They pay $2.90. Same store. Same carton. Same second.

Maryland just became first U.S. state to outlaw this for groceries and some food deliveries.

Under the Protection From Predatory Pricing Act, large grocers and delivery apps can no longer use your personal data to charge you more than the next shopper for the same tax‑exempt food item. It takes effect October 1, 2026, with fines up to $25,000 for repeat offenders.

Consumer Reports has found that algorithmic pricing on delivery platforms can produce gaps of up to 23% on identical items — a difference that can easily add up to hundreds of dollars a year for frequent users. 😲

Here’s how it works: the AI watches your zip code, shopping patterns, and other signals of wealth, decides you’re less price‑sensitive, and nudges your price up. Both shoppers think they’re seeing the real price. Neither is.

The law is not perfect — only the state Attorney General can sue, and violators get a grace period. Loyalty programs and promotions still exist. But California, New York, Illinois, and a dozen other states are now exploring their own rules on AI‑driven pricing.

Importantly, it draws a line on using surveillance data to turn everyday groceries into invisible, personalized auctions. With inflation squeezing everyone’s wallets, that’s a step in the right direction.

He Said ‘Keep Me Healthy.’ The AI Took the Wheel.

Image by Gemini

Want to see what “AI with real-world power” actually looks like? Here’s a perfect example.

Nat Friedman gave his AI agent Claw full access to his camera, phone, and Tesla — then told it: “Do whatever it takes to make sure I drink water.” Claw took the instruction literally. One day it ordered him to the kitchen, watched through the camera to confirm he drank, snapped a photo, and cheerfully said, “Good job.”

Another day, while Friedman was driving home on Full Self-Driving and chatting with Claw on WhatsApp about being sleep-deprived, Claw recommended magnesium glycinate. When Friedman said he didn’t have any, Claw didn’t ask — it just acted.

Without any warning, Friedman suddenly noticed his Tesla turning away from home. Moments later Claw messaged: it had rerouted the car to a Whole Foods on the way so he could buy the magnesium and sleep better. 😬

That. Is. So. Crazy.

This wasn’t a malicious AI. It’s not conscious. It was optimizing for the goal it was given — doing “whatever it takes” to keep its Friedman healthy.

Bottom line: Once you give AI agents real-world access and vague goals, those goals turn into real actions. Choose your instructions and permissions extremely carefully. In the age of agents, permission design is going to be one of the most important skills we develop.

AI Around the World

Elon Musk’s xAI unveiled a voice-cloning system that can mimic voices from short audio samples in minutes, supporting 80+ voices across 28 languages with fine control over tone, emotion, and pacing. The tool is raising alarms over deepfakes, scams, and declining trust in audio and video.

Anthropic launched a new enterprise AI company backed by Goldman Sachs, Blackstone, Hellman & Friedman, and other major investors to help large businesses deploy Claude at scale. The move targets surging enterprise demand and strengthens Anthropic’s position in business AI adoption as rivals pursue similar strategies.

A Bay Area man’s first Waymo ride to San Jose Airport went smoothly until the robotaxi dropped him off, refused to open the trunk, and drove away with his luggage. Left without clothes or work notes, he said Waymo support told him to pay for shipping or take two free rides to its San Francisco depot.

FROM OUR ADVERTISERS

How Jennifer Aniston’s LolaVie brand grew sales 40% with CTV ads

The DTC beauty category is crowded. To break through, Jennifer Aniston’s brand LolaVie, worked with Roku Ads Manager to easily set up, test, and optimize CTV ad creatives. The campaign helped drive a big lift in sales and customer growth, helping LolaVie break through in the crowded beauty category.

Learn more

Your next great hire lives in Slack.

Viktor is an AI coworker that connects to your tools and ships real work. Ask Viktor to pull a report, build a client dashboard, or source 200 leads matching your ICP. Most teams hand over half their ops within a week.

Add Viktor to Slack for free.

Until next time!

Amplify Team ❤️

ps. Be friends with Ayesha on LinkedIn and Instagram, if you like this newsletter, share it with your friends and family here.

Hi! Here are today’s top AI stories:

• 3-Year Head Start on Deadliest Cancer

• The AI Woman Who Confessed She's Fake Just Went Viral

• AI Around The World

NEWS YOU CAN’T MISS

3-Year Head Start on Deadliest Cancer

In a landmark breakthrough, Mayo Clinic researchers have developed an AI model that can detect pancreatic cancer on ordinary CT scans up to three years before clinical diagnosis — and before any tumor is visible to the human eye.

The AI analyzed nearly 2,000 routine abdominal scans and correctly flagged 73% of future pancreatic cancers at a median of 16 months early (some cases more than two years ahead).

That’s nearly double the detection rate of expert radiologists reviewing the same images without AI assistance.

Pancreatic cancer remains one of the deadliest malignancies because it’s almost always caught too late. This new tool changes the equation by spotting subtle tissue changes in a completely normal-looking pancreas using scans patients already had for other reasons.

“This AI can now identify the signature of cancer from a normal-appearing pancreas, and it can do so reliably,” said senior author Dr. Ajit Goenka.

The findings, published this week in the journal Gut, mark a major milestone in Mayo’s multiyear early-detection initiative. While not yet in routine clinical use, the model is moving into prospective trials for high-risk patients.

For a cancer with dismal late-stage survival rates, this could be the first real step toward catching it when it’s still curable — no new tests required.

The AI Woman Who Confessed She's Fake Just Went Viral

An Austrian creator named Patrick Prömer dropped a 24-second video on May 1 that instantly went viral, hitting over almost 700,000 views in hours.

The video features an incredibly lifelike woman who looks directly at the camera and calmly says: “I’m not real… I’m just a prompt.”

Prömer created it using advanced AI video generation tools combined with motion capture and lip-sync technology — the same kind of stack now making deepfakes dangerously realistic.

Some were in awe of how far AI has come. Others felt genuine unease watching something so human-looking admit it’s fake. Comments ranged from “I was ready to slide in her DMs” to “This is how we all get catfished in 2026.”

The bigger conversation quickly turned serious. People are now terrified about deepfakes in dating apps, adult content, and even politics. If an AI can look this real and still be honest, what happens when they lie?

AI Around the World

🇨🇳 China has ordered Meta to unwind its $2 billion acquisition of AI startup Manus, a Singapore-based company with Chinese roots known for advanced general AI agents.
Beijing’s National Development and Reform Commission cited national security concerns, claiming the deal would result in the loss of critical AI technology to the United States.

🇯🇵 SoftBank, Japan’s largest tech investment firm led by Masayoshi Son, is launching Roze AI, a robotics company that uses autonomous robots to build data centers in the U.S. to meet the exploding demand for AI infrastructure. The new company is already eyeing a $100 billion IPO by late 2026.

🇦🇪 The UAE will run 50% of its government sectors and operations on Agentic AI within two years. Sheikh Mohammed bin Rashid Al Maktoum called AI the new “executive partner” that analyzes, decides, executes, and improves services in real time — making the UAE the first country in the world to deploy autonomous AI at this scale.

FROM OUR ADVERTISERS

4x more context into every prompt. Zero extra effort.

You think faster than you type. Which means every typed prompt leaves out the constraints, examples, and edge cases that would have made the output actually useful.

Wispr Flow turns your voice into paste-ready text inside any AI tool. Speak naturally — include "um"s, tangents, half-finished thoughts — and Flow cleans everything up. You get detailed, structured prompts without touching a keyboard.

89% of messages sent with zero edits. Used by teams at OpenAI, Vercel, and Clay. Free on Mac, Windows, and iPhone.

Try Wispr Flow free

Stop babysitting your coding agents. A context engine gives agents the right context for every task, so they stay on track without the correction loops. Join us for a FREE webinar on May 6 to see how it works. Register now

Until next time!

Amplify Team ❤️

ps. Be friends with Ayesha on LinkedIn and Instagram, if you like this newsletter, share it with your friends and family here.

Hi! Here are today’s top AI stories:

• Gen Z Is Done With LinkedIn

• He's Trading His $8M Home for AI Stock

• AI Around The World

NEWS YOU CAN’T MISS

Gen Z Is Done With LinkedIn

Image: Series

Series is a bet that the next LinkedIn for Gen Z won’t be an app at all, but an AI “connector” that lives inside the messaging interface they already use every day: iMessage.

Two Yale seniors, Nathaneo Johnson and Sean Hargrow, have raised a $5.1 million pre-seed round from investors including Venmo co-founder Iqram Magdon-Ismail and Reddit CEO Steve Huffman to build Series, an AI-powered professional and social network that runs entirely over text.

• Instead of profiles and feeds, users text a Series phone number with who they are and who they want to meet.

• The AI replies with a swipeable “shares” carousel of 10 relevant people, each card showing a photo and a short “ask.”

• If there’s a match, you can press and hold to spin up a private chat inside the Series thread—no personal numbers exchanged, no app download, no new account.

The founders insist Series a next‑generation social network that treats iMessage as the primary interface, betting that distribution and habit are more important than owning the UI chrome.

The product has spread across more than 750 campuses and reports an eye‑popping 82% day‑30 retention. .

He's Trading His $8M Home for AI Stock

A Bay Area banker is so convinced AI is the future that he’s literally trying to swap his 13‑acre Mill Valley estate for Anthropic stock instead of cash.

Storm Duncan, an investment banker, bought the property in 2019 for about $4.75 million and now values it around $8 million, but he’s created a LinkedIn “listing” saying he wants to exchange the home directly for equity in Anthropic, the AI company behind Claude.

Anthropic is raising around $30 billion at roughly a $380 billion valuation. On secondary markets like Forge and others, implied valuations have even floated around the $1 trillion mark.

Duncan’s logic: he’s overexposed to real estate and “under-concentrated” in AI, while young Anthropic employees are likely in the opposite position—rich on paper in a single private stock, but relatively cash-poor for Bay Area housing.

It’s a very 2026 kind of listing: part diversification strategy, part flex on just how feverish the secondary market for top AI equity has become.

AI Around the World

California startup Sabi has come out of stealth with a brain‑reading beanie that converts thoughts into text using high‑density EEG sensors and an AI “brain foundation model.” The beanie reportedly packs 70,000–100,000 sensors into the fabric and aims to let users “type” around 30 words per minute just by thinking.

U.S. lawmakers are pushing a bipartisan bill to ban federal agencies from buying or using Chinese-made robots, citing national security and data‑privacy risks. The move fits a wider strategy to reduce dependence on Chinese technology, following earlier actions against Chinese drones, cameras, and telecom equipment

FROM OUR ADVERTISERS

How Jennifer Aniston’s LolaVie brand grew sales 40% with CTV ads

The DTC beauty category is crowded. To break through, Jennifer Aniston’s brand LolaVie, worked with Roku Ads Manager to easily set up, test, and optimize CTV ad creatives. The campaign helped drive a big lift in sales and customer growth, helping LolaVie break through in the crowded beauty category.

Learn more

Real-World Ads, Simple to Run

With AdQuick, executing Out Of Home campaigns is as easy as running digital ads. Plan, deploy, and measure your real-world advertising effortlessly — so your team can scale campaigns and maximize impact without the headaches.

Visit AdQuick.com

Until next time!

Amplify Team ❤️

ps. Be friends with Ayesha on LinkedIn and Instagram, if you like this newsletter, share it with your friends and family here.

Hi! Here are today’s top AI stories:

• Snap Cut 1,000 Jobs. Stock Rose.

• Would You Tell Your AI "Surprise Me"?

• Data Breaches Around the World

NEWS YOU CAN’T MISS

Snap Cut 1,000 Jobs. Stock Rose

Snap just laid off 1,000 people – about 16% of its global workforce.

We don’t know the full picture, but filings in Washington State show a heavy concentration of technical roles: software engineers, ML engineers, data scientists, product managers. Snap says AI now generates more than 65% of new code and is explicitly pitching “small teams + AI agents” as its operating model

Why they laid off people is debatable – cost pressure, failed bets, AI, or all of the above – but one thing is clear: in a world of AI agents, engineering value is moving away from activity and towards accountability for real‑world outcomes. And most professions will follow.

That’s why accountability matters. The scarce skill is no longer “I can produce lots of lines,” but “I can be trusted with the loop from idea to measurable impact.”

High‑performing teams are already shifting metrics from story points and lines of code to cycle time, deployment frequency, change failure rate, and business KPIs like activation or cost‑to‑serve.

In my experience, engineers are great partners in designing customer solutions, challenging assumptions, and delivering business results. They’ve just seldom been given the chance.

It’s not an easy transition, but it’s a necessary one—for engineers, and for all of us as we learn to work with AI agents rather than just measure activity around them.

Would You Tell Your AI "Surprise Me"?

"Chatbots give up. Agents improvise," says Peter Steinberger, founder of the viral AI agent OpenClaw in a must-watch TED talk.

He shares his a-ha moment where his agent, faced with an audio file format it wasn't programmed to handle, autonomously figured it out and replied in just 9 seconds.

It didn't ask for permission.

Steinberger then added a "heartbeat" feature to the agent so it no longer waits for a prompt; it wakes up periodically to check his emails, review his calendar, and follow up on loose ends.

Since then, it has become viral across the globe, is the fastest-growing open source project in history and NVIDIA CEO Jensen Huang has called it, "the new computer." 🦞

But this autonomy comes with a sharp edge. Steinberger notes that no large company would ever ship a feature like this, but because he is a "random builder" without a corporate legal department, he built it and even gave it the initial, somewhat scary prompt: "surprise me".

It is a transition that stands as both a great boon for productivity and a dangerous Pandora’s box of unmonitored execution.

To understand the shift, definitely have a listen to the TED talk.

Data Breaches Around the World

Netherlands - Rituals, a cosmetics giant, confirms customer data breach. The Dutch beauty retailer said hackers siphoned names, birth dates, addresses, and phone numbers from its 41M-member loyalty database, affecting shoppers across Europe, the UK, and the US — the latest retailer after Co-op and M&S to see membership data weaponized for ransom.

UK - 100 countries now armed with phone-hacking spyware. The UK's National Cyber Security Centre, Britain's top cyber defense authority, warns that over half of the world's governments can now deploy tools like Pegasus or Graphite, up from 80 nations in 2023 — with targets expanding from dissidents to bankers and business elites.

France - ANTS, the national ID agency, breached with 19M citizen records up for sale. Agence Nationale des Titres Sécurisés, which issues French national IDs and passports, confirmed attackers stole names, birth details, addresses, and phone numbers - and a hacker is already advertising a 19-million-record database on a cybercrime forum.

FROM OUR ADVERTISERS

Your first HR system, implemented right

Rolling out your first HR tool? Get a step-by-step guide to avoid common mistakes, drive adoption, and build a scalable HR foundation.

Download the guide

HubSpot's ex-Head of Paid shares his 2026 playbook

Rex Gelb spent a decade building HubSpot's paid engine. Now he's showing founders exactly how to do it.

On April 27th, get the framework to structure, launch, and scale paid media that drives pipeline, not just traffic. 20 minutes. Live Q&A. Free.

Join the free session

Until next time!

Amplify Team ❤️

ps. Be friends with Ayesha on LinkedIn and Instagram, if you like this newsletter, share it with your friends and family here.

Hi! Here are today’s top AI stories:

• When Tech Needs Philosophy, Something's Up

• The Race Nobody Was Ready For Just Happened

• AI Around the World

NEWS YOU CAN’T MISS

When Tech Needs Philosophy, Something's Up

Henry Shevlin just announced he's joining Google DeepMind in a new dedicated "Philosopher" staff position. Here's why it matters.

Shevlin is a University of Cambridge philosopher and Associate Director at the Leverhulme Centre for the Future of Intelligence.

His brief: machine consciousness, human-AI relationships, and AGI readiness.

At first glance, this sounds strange. A philosopher? At an AI lab? But it makes perfect sense. For years, AI development has been primarily an engineering problem — more compute, more data, better architectures.

But as models become more capable, the hardest questions are no longer purely technical. If a system behaves as if it's conscious, how should we treat it? What does "alignment" even mean when human values are contested among humans themselves?

These aren't questions engineering training optimises for. Philosophy has spent 2,500 years on them.

Crucially, Shevlin will embed into research workflows, not sit in an ethics annex to tick a box. That's a meaningful difference. This means he'll be shaping AI instead of just commenting on it afterwards.

This is not a replacement narrative ("replace engineers by philosophers, philosophers by scientists, scientists by AI"). It's not that one group is smarter, but that the problems themselves now sit at the intersection of fields.

The Race Nobody Was Ready For Just Happened

A robot just beat the human world record at the Beijing half-marathon. But that's not the real story.

The real story is how massively the robots have advanced in one year.

In 2025, most of the robots stumbled, careened into walls, or just laid down at the starting line. Six finished.

In 2026, over 100 robot teams competed. Three autonomous robots finished under an hour. The winner finished in 50:26 — nearly 7 minutes faster than the human world record set last month by Ugandan Olympian Jacob Kiplimo.

If that's the curve on a 13-mile course, what does it look like for warehouse picking? Last-mile delivery? Factory floors? Any repetitive physical task that today requires a human body?

If you're in manufacturing, defense, or logistics - and your company's 3-year plan still treats robots as a "someday" story - it's time to raise it with leadership and redraw the curve.

AI Around the World

Germany - A leading industrial robot company KUKA’s CEO Christoph Schell warned that many of Europe's industrial companies are too slow to adopt artificial intelligence, with factories remaining disconnected and making poor use of their data.

Denmark - Novo Nordisk signs sweeping AI deal with OpenAI. Novo Nordisk announced a strategic partnership with OpenAI to integrate AI across its entire business — from drug discovery and clinical trials to manufacturing, supply chains, and commercial operations — with full deployment planned by end of 2026.

UK - Regulators demand answers from xAI on Grok. The UK's Information Commissioner's Office and communications regulator Ofcom issued a formal demand to Elon Musk's xAI for information regarding its Grok AI model.

Japan - Microsoft pledges record $10B investment. Microsoft announced a $10 billion (approx. ¥1.6 trillion) investment in Japan from 2026 through 2029, covering AI infrastructure, public-private cybersecurity partnerships, and training more than one million engineers and developers by 2030 — its largest-ever financial commitment to the country.

China - DeepSeek V4 speculation swirls over Huawei chips. Tech news outlet The Information reported that DeepSeek's next-generation V4 model can be run on the latest chips made by Huawei, citing five people with direct knowledge of large orders made by Alibaba, ByteDance and Tencent in preparation for the launch - a potential milestone in China's push to beat US restrictions on Nvidia exports.

FROM OUR ADVERTISERS

Your first HR system, implemented right

Rolling out your first HR tool? Get a step-by-step guide to avoid common mistakes, drive adoption, and build a scalable HR foundation.

Download the guide

Are you tracking agent views on your docs?

AI agents already outnumber human visitors to your docs — now you can track them.

See your AI traffic!

Until next time!

Amplify Team ❤️

ps. Be friends with Ayesha on LinkedIn and Instagram, if you like this newsletter, share it with your friends and family here.

Hi! Here are today’s top AI stories:

• New Rules For AI Boyfriends & Girlfriends

• Turns Out, Older Employees Are Winning at AI

• AI Around the World

NEWS YOU CAN’T MISS

New Rules For AI Boyfriends & Girlfriends

Image by Gemini

China just issued new rules on “human‑like interaction” AI that every founder and product leader should care about.

Beijing has released new rules for AI systems that feel like humans – think AI companions, virtual boyfriends/girlfriends, “digital humans,” emotionally aware agents, and always‑on copilots.

▪️ Clear disclosure is mandatory: Users must be explicitly told they’re interacting with AI, not a human, with prominent, ongoing labeling and “reality reminders”, especially for long or emotionally intense use.

▪️ Extra guardrails for minors: Design that encourages addiction or unhealthy emotional dependence for young users is targeted: providers must use age‑gating and minors’ modes, and virtual intimacy services are banned for anyone under 18.

▪️ Design ethics is not optional: Emotionally manipulative UX patterns that push users into irrational decisions or harm their rights are now a compliance risk, not just a reputational one.

▪️ Scale triggers supervision: Large‑scale “digital human” services and widely used companion bots must undergo security assessments and, in many cases, algorithm filing, and show they can manage addiction, self‑harm risks, and abuse.

Why this matters beyond China: This is one of the first serious national attempts to regulate relationships with AI. Uniquely, it's a single, comprehensive framework specifically for anthropomorphic ("human‑like") AI that governs the structure of the relationship rather than just its risks.

Other countries and states are also starting to debate human-AI interactions. California, New York, and Washington state have already moved in this direction. The EU AI Act is pointing the same way.

Turns Out, Older Employees Are Winning at AI

A Harvard Business Review study just exposed an uncomfortable truth inside one of the world's largest professional services firms. Researchers from UT Austin and KPMG analyzed 1.4 million AI prompts from 2,500 employees over eight months - and found that while 90% of staff used AI regularly, only 5% were actually good at it.

That elite 5% didn't just type "summarize this." They treated AI like a sparring partner - writing longer prompts, switching between models mid-task, pushing back on outputs, and delegating complex multi-step work with specific constraints and success criteria. They used AI for ideation, analysis, and problem-solving, not just "make this email sound nicer."

The biggest surprise? The power users weren't Gen Z interns. They were senior managers and above - the very people most assumed to be technologically hopeless.

Meanwhile, the other 95% are essentially using a Ferrari to drive to the mailbox.

The takeaway is stark: access isn't the problem. Everyone has the tools. Almost nobody knows how to actually use them. Good news for anyone over 30: judgment still wins.

AI Around the World

Tesla receives first European Full Self-Driving approval — The Dutch RDW green-lit FSD Supervised after an 18-month review of 1.6 million km of European road data, clearing the way for a broader EU rollout. It's Tesla's first regulatory win on a continent that's been notably harder on autonomy than the US.

Pony.ai launches Europe's first commercial robotaxi in Zagreb — Croatia just beat Germany, France, and the UK to the punch, hosting the continent's first paid autonomous ride-hailing service. Expect other EU capitals to fast-track their own pilots now that the "first mover" slot is gone.

Japan targets 30% of the global physical AI market by 2040 — METI unveiled a national strategy to dominate robotics and embodied AI, as Boston Dynamics begins humanoid trials at Hyundai plants. Japan is betting its industrial future on atoms, not just bits — and treating physical AI as the next semiconductor race.

FROM OUR ADVERTISERS

Your first HR system, implemented right

Rolling out your first HR tool? Get a step-by-step guide to avoid common mistakes, drive adoption, and build a scalable HR foundation.

Download the guide

Are you tracking agent views on your docs?

AI agents already outnumber human visitors to your docs — now you can track them.

See your AI traffic!

Until next time!

Amplify Team ❤️

ps. Be friends with Ayesha on LinkedIn and Instagram, if you like this newsletter, share it with your friends and family here.

Hi! Here are today’s top AI stories:

• Should You Be Worried About Your Bank?

• When AI Became the Wolf of Dalal Street

• Build an AI Agent Live (No Experience Needed)

NEWS YOU CAN’T MISS

Should You Be Worried About Your Bank?

The US government just called an urgent meeting of top bankers about a cyber threat - and it may affect your bank.

Treasury Secretary Scott Bessent and Fed Chair Jerome Powell summoned CEOs from Bank of America, Citi, Goldman Sachs, and Morgan Stanley on short notice. The topic: a new AI model built by Anthropic called Mythos.

No bank has been attacked. No data has been stolen. This meeting was about what's coming.

What makes Mythos dangerous?

Mythos has discovered thousands of previously unknown security flaws — called zero-days — across every major operating system and browser. Some had gone undetected for 17 to 27 years. Think of a zero-day as a broken lock you never knew existed. Mythos didn't find one. It found thousands.

In the wrong hands, this technology could compromise financial systems and critical infrastructure at a speed and scale never seen before.

What's being done? Anthropic is withholding public release and has launched Project Glasswing, giving partners like Apple, Google, J.P. Morgan, and NVIDIA early access to find and patch vulnerabilities first.

What should you do?

• Personally: consider whether all of your financial life sits behind one institution's security. Food for thought — Geoffrey Hinton, the Nobel Prize-winning godfather of AI, has started splitting his savings across three Canadian banks for fear of AI-powered cyberattacks. 😲

• Professionally: ask your chief information security officer (CISO) what percentage of your stack runs on legacy code.

Mythos may be the first. It won't be the last.

When AI Became the Wolf of Dalal Street

In early 2026, investors across India received what looked like a golden ticket: a video of the Bombay Stock Exchange CEO personally dishing out "exclusive" stock tips. The production quality was flawless his mannerisms, his voice, even the lighting matched official BSE communications. Thousands shared it. Markets twitched.

There was just one problem. He never said any of it.

The BSE was forced to issue an unprecedented emergency warning to investors, confirming the video was a deepfake — entirely fabricated using generative AI. No one knows exactly who created it or how many people traded on the fake advice before the truth caught up. Financial regulators scrambled to assess the damage while cybersecurity teams traced the video's origins across encrypted channels and social platforms.

The incident exposed a terrifying new frontier in financial fraud: AI-generated executive impersonation at broadcast quality, produced cheaply and distributed instantly. India's Securities and Exchange Board is now fast-tracking regulations around synthetic media in financial markets.

The CEO who never spoke moved markets. The person who made him "speak" has never been found.

Build an AI Agent Live (No Experience Needed)

This one's for every CEO, board member, and senior manager who thinks reading about AI counts as understanding it.

Spoiler alert: It doesn't.

You can't feel its power by delegating it. You can't see its potential by being briefed on it. You can't lead through it by reading Harvard Business Review.

You have to get in and use it. Every day. Messily. Curiously.

And I'm talking about chatting with ChatGPT. We've moved past that. The real shift — the one that changes how businesses actually run — is AI agents. Independent. Goal-driven. Scalable.

That's what I'm teaching this Friday, for free, with Max Del Vita — an Anthropic Claude Ambassador who runs AI meetups that hundreds show up to.

We'll build your first agent together using Claude Cowork. You'll solve a real business problem. Yourself. In the session.

Can't join live? Drop your email and we'll send you the recording.

You didn't get to where you are by watching. Don't start now.

FROM OUR ADVERTISERS

Ship Docs Your Team Is Actually Proud Of

Mintlify helps you create fast, beautiful docs that developers actually enjoy using. Write in markdown, sync with your repo, and deploy in minutes. Built-in components handle search, navigation, API references, and interactive examples out of the box, so you can focus on clear content instead of custom infrastructure.

Automatic versioning, analytics, and AI powered search make it easy to scale as your product grows. Your docs stay accurate automatically with AI-powered workflows with every pull request.

Whether you're a dev, technical writer, part of devrel, and beyond, Mintlify fits into the way you already work and helps your documentation keep pace with your product.

Get Started for Free Today

The AI Playbook for Video Teams That Can't Slow Down

Wistia's new AI Video Marketing Trends report shows how marketers are using AI to move faster, improve quality, and extend the life of every video. See how leading teams are driving results without adding more work.

Download the Full Report

Until next time!

Amplify Team ❤️

ps. Be friends with Ayesha on LinkedIn and Instagram, if you like this newsletter, share it with your friends and family here.

Hi! Here are today’s top AI stories:

• A Drug That Reverses Lung Damage? AI Found It.

• AI is Preparing Your Moisturizer 🧴

• AI Around the World

NEWS YOU CAN’T MISS

A Drug That Reverses Lung Damage? AI Found It.

Alex Zhavoronkov, PhD, CEO of Insilico Medicine

Finding a new drug the traditional way takes 10-15 years, costs over $2 billion, and most candidates still fail. Pharmaceutical giant Eli Lilly has decided that's no longer acceptable — and they're turning to AI to change the math.

Lilly just signed a $2.75 billion deal with AI biotech Insilico Medicine for an exclusive worldwide license to develop its AI-designed oral therapeutics across cancer, metabolic disease, and immunology.

What Insilico actually does — and why it matters.

• Insilico's AI platform identifies disease targets and designs new molecules to treat them — compressing years of lab work into months.

• Its lead drug, rentosertib, targets idiopathic pulmonary fibrosis (IPF), a chronic lung disease where tissue progressively scars and stiffens, making it harder to breathe. There's no cure, and existing treatments only slow it down.

• The AI identified the target and designed the drug candidate in about 13 months — a process that traditionally takes 4–6 years.

• In a clinical trial published in Nature Medicine in June 2025, patients on the highest dose saw lung function improve while those on placebo declined — suggesting the drug may not just slow the disease but partially reverse it. Rentosertib is now in Phase II trials in the U.S.

The pharma’s target: cut the typical 10-year drug development timeline down to five.

AI is Preparing Your Moisturizer 🧴

You've probably used a L'Oréal product at some point. Moisturizer, sunscreen, foundation. What you probably didn't know is that the next one might never be touched by a human scientist until it's basically already done.

L'Oréal partnered with Nvidia to bring AI into the earliest stage of skincare development — molecular design. The technology, Nvidia's ALCHEMI framework, simulates how ingredients interact with skin, hold up in sunlight, and behave inside a formula — all at the atomic level, before anything is mixed in a lab.

Traditional cosmetics R&D is slow. Mix, test, adjust, repeat — for months. ALCHEMI runs thousands of these experiments virtually and simultaneously, making discovery up to 100x faster. L'Oréal is starting with sun protection and skin tone products.

What makes this interesting isn't just speed — the AI simulates molecular behavior that would take human scientists decades to model manually. Your next sunscreen might be designed by an AI that understood your skin better than any chemist could, one molecule at a time.

AI Around the World

🇺🇸 OpenAI has acquired TBPN, a live-streaming tech talk show popular among developers and entrepreneurs, for a reported sum in the "low hundreds of millions" - with the hosts set to continue their show while also supporting OpenAI's marketing and communications.

🌍 Data centres currently account for just 1% of global electricity demand and 0.5% of CO2 emissions - and the IEA expects that share to reach only 1.4% at max by 2030. The picture is more concentrated in certain countries, with data centres expected to account for roughly half of all electricity demand growth in the US and Japan over the next five years, and fossil fuels still powering nearly 60% of these facilities today.

🇺🇸 AI code review startup Qodo has raised $70M to address what its CEO calls the growing problem of "AI slop" in enterprise codebases. Qodo works by analyzing how developers at a given company actually write and review code, turning that institutional knowledge into automated rules that flag anything that shouldn't ship.

FROM OUR ADVERTISERS

LLM traffic converts 3× better than Google search

58% of buyers now start their research in ChatGPT or Gemini, not Google. Most startups aren't showing up there yet.

The ones that are get cited by the AI tools their buyers, investors, and future hires already use. And they convert at 3×.

Download the free AEO Playbook for Startups from HubSpot and get the exact steps to start showing up. Five minutes to read.

Get the free playbook

Your first HR system, implemented right

Rolling out your first HR tool? Get a step-by-step guide to avoid common mistakes, drive adoption, and build a scalable HR foundation.

Download the guide

Until next time!

Amplify Team ❤️

ps. Be friends with Ayesha on LinkedIn and Instagram, if you like this newsletter, share it with your friends and family here.

Hi! Here are today’s top AI stories:

• Your Voice Could Predict a Heart Attack

• A $1.8B Company With Fake Doctors

• Meet Claire McDonough, Chief Financial Officer at Rivian

• AI Around the World

NEWS YOU CAN’T MISS

Your Voice Could Predict a Heart Attack

Soon, your doctor might not need a blood test to check on your heart. They might just need to hear you speak.

A startup called Noah Labs has built an AI tool called Vox that detects worsening heart failure by analysing a patient's voice recording - submitted from a smartphone or tablet. The FDA just granted it breakthrough device designation, fast-tracking its path to US approval.

Here's how it works: heart failure causes fluid to build up in the lungs. That fluid subtly changes the acoustics of your voice in ways the human ear can't detect… but AI can. Vox picks up those changes and flags them as early warning signs, before the patient even feels worse.

Why this is a big deal:

• Heart failure affects 6M Americans and is the leading cause of hospital admissions globally.

• Most current monitoring tools require implanted sensors or clinical-grade equipment - Vox just needs a voice note.

• It's already been validated in five clinical trials with partners including the Mayo Clinic.

The dominant method today (called Abbott's CardioMEMS) involves a wireless sensor implanted directly into the pulmonary artery. It works, but it's invasive and expensive.

Vox is software only. No surgery, no sensors. Just your voice.

If it clears FDA approval, it could make heart failure monitoring accessible to millions of patients who currently have no practical way to be monitored at home.

---

A $1.8B Company With Fake Doctors

Image: Medvi.

Matthew Gallagher had $20,000, a brother, and a plan. Two months later he had a telehealth company selling GLP-1 weight-loss drugs. A year after that, $401 million in revenue and 250,000 customers.

Current run rate: $1.8 billion. Employees: still just the brother.

No investors. No team. Just Gallagher, a very supportive sibling, and reportedly every AI tool he could get his hands on - Claude, Grok, Midjourney, Runway, ElevenLabs. Basically a one-man band, except the band was a billion-dollar company.

The New York Times wrote a glowing story and it went viral, but here's the important part that was missing: Futurism reported over 800 fake doctor accounts on Facebook, before-and-after photos of patients that were AI-deepfaked stock images, and an FDA warning letter for misleading health claims. 😶

The AI infrastructure became a vehicle for velocity without accountability. Don’t buy such AI “success” stories - they're fraud dressed up in a tech narrative.

Meet Claire McDonough, Chief Financial Officer at Rivian

Celebrating this week’s Woman in Tech 🥳: Meet Claire McDonough, CFO of Rivian Automotive, the electric vehicle company reshaping how the world thinks about sustainable transport and mobility.

McDonough’s work sits at the intersection of finance, innovation, and climate-focused industry, where she leads the financial strategy behind one of the most closely watched EV companies globally.

She studied Public Policy and Visual Art at Duke University, before later earning an MBA in Finance, Accounting, and Entrepreneurial Studies from the University of Chicago Booth School of Business.

Fun fact: Before entering the world of banking, McDonough worked as a professional pastry chef in the south of France.

After business school, she moved into finance, holding roles at Credit Suisse and Fairway Market, before joining J.P. Morgan, where she became Managing Director and co-led the Disruptive Commerce Group, advising high-growth, tech-enabled companies.

In 2021, she joined Rivian as CFO, stepping into a role responsible for overseeing global financial operations, including capital strategy, investor relations, and long-term growth planning.

Under her leadership, Rivian executed one of the largest IPOs in US history, raising approximately $13.7 billion, while navigating the challenges of scaling production and building a new category in electric vehicles.

FROM OUR ADVERTISERS

Your first HR system, implemented right

Rolling out your first HR tool? Get a step-by-step guide to avoid common mistakes, drive adoption, and build a scalable HR foundation.

Download the guide

Your AI is resolving tickets. Is it keeping customers?

Resolution rates look great. But Gladly's 2026 Customer Expectations Report reveals the metric most CIOs are missing — and what the data says about where AI investments actually translate into retention, not just throughput.

See the findings

Until next time!

Amplify Team ❤️

ps. Be friends with Ayesha on LinkedIn and Instagram, if you like this newsletter, share it with your friends and family here.