Good morning. Silicon Valley is queuing up Theranos II as Billy Evans – Elizabeth Holmes’ husband – works the Sand Hill circuit for $50 million to fund Haemanthus, a laser-powered blood-tester that pricks pets today and people tomorrow. Investors, kindly place your cash (and skepticism) in the centrifuge before proceeding.

— The Deep View Crew

In today’s newsletter:

• 🏥 AI for Good: A selfie health assessment

• 💰 OpenAI negotiates with Microsoft

• 🤖 Amazon’s Vulcan learns to feel

🏥 AI for Good: A selfie health assessment

Hospitals already snap a photo of you when you arrive, now FaceAge turns that into new data for doctors. Trained on more than 58,000 “healthy” faces and tested on 6,000 cancer patients, the deep-learning system estimates biological age – how old a body seems, not how many birthdays it’s had. Across three independent cancer cohorts, every extra decade FaceAge estimated in age raised the risk of death 11-15%. In palliative patients, plugging FaceAge into a standard survival model lifted predictive accuracy from 0.74 to 0.80 AUC (Area Under the ROC Curve), roughly the jump clinicians get from a CT scan to a PET scan.

Why it matters: Oncologists often rely on gut feeling to judge whether a patient is healthy enough for chemotherapy. FaceAge brings objectivity to that judgment, turning a patient’s appearance into a quantifiable, clinically useful score.

How it works:

• Face Detection: A neural network identifies and processes the face from a standard photograph.

• Biological Age Estimation: An Inception-ResNet model predicts a patient’s biological age with a 4-year margin of error for seniors.

• Clinical Forecasting: The score feeds existing risk tools, boosting accuracy by up to six percentage points.

• Genetic Correlation: It also tracks closely with senescence-related genes, suggesting it taps into real molecular aging processes.

Zoom in: Training images lean toward public figures, so broader, representative datasets are essential to curb demographic bias. A prognostic tool that guides therapy must clear rigorous validation and full transparency. Without firm policy, insurers or employers could weaponize “age-in-face” scores; guardrails need to be in place before deployment.

The big picture: FaceAge shows how AI can repurpose everyday data for health – what heart-rate sensors did for watches, facial analytics might do for cameras. If larger trials confirm the findings and equity concerns are addressed, the next time you visit the doctor, biological age may be a standard metric produced as part of your checkup “chronological age: 51; biological (FaceAge): 38”, and treatment plans might shift accordingly.

✂️ Cut your QA cycles down from hours to minutes

If slow QA processes and flaky tests are a bottleneck for your engineering team, you need QA Wolf.

QA Wolf's AI-native platform supports both web and mobile apps, delivering 80% automated test coverage in weeks and helping teams ship 5x faster by reducing QA cycles to minutes.

With QA Wolf, you get:

✅ Unlimited parallel test runs

✅ 15-min QA cycles

✅ 24-hour maintenance and on-demand test creation

✅ Zero-flake guarantee

The result? Drata’s team of 80+ engineers saw 4x more test cases and 86% faster QA cycles.

No flakes, no delays, just better QA — that’s QA Wolf.

Schedule a demo to learn more

💰 OpenAI negotiates with Microsoft

OpenAI and Microsoft are in tense negotiations to rewrite the terms of their multibillion-dollar AI alliance, aiming to clear the way for a future OpenAI IPO while securing Microsoft long-term access to OpenAI’s advanced models. A key sticking point is Microsoft’s stake: the tech giant has poured $13 billion into OpenAI and is even willing to reduce its equity share in exchange for guaranteed access to OpenAI’s innovations beyond 2030. OpenAI, for its part, plans to convert its for-profit arm into a public benefit corporation to attract billions in fresh funding for ever more powerful AI – a structural shift that still needs Microsoft’s blessing as the startup’s largest backer.

The details:

•  OpenAI has already signaled it will halve Microsoft’s cut of future revenues (to 10% from 20% by 2030)

• OpenAI will likely convert its business arm into a public benefit corporation, allowing outside investment while keeping its nonprofit mission intact.

• The friction comes as OpenAI ramps up competition, targeting enterprise clients and planning massive computer projects with Oracle and SoftBank.

Microsoft’s insiders have bristled at OpenAI’s newfound assertiveness – one executive reportedly slammed the startup’s “bad partner attitude,” saying OpenAI expects Redmond to “give us money and compute and stay out of the way”.

What’s next: Despite the strain, the partnership isn’t breaking. OpenAI’s services still run primarily on Microsoft’s Azure cloud, and its models power flagship products like Bing Chat and GitHub’s Copilot. Every ChatGPT query still routes through Azure, and Microsoft continues to integrate OpenAI’s tech across its lineup.

The big picture: As OpenAI pushes toward IPO readiness, the delicate balance between mission, money, and control is under pressure. Microsoft may end up with a smaller slice of equity but a deeper lock-in on tech access. Both sides are betting this restructure won’t just preserve the alliance but future-proof it.

Is the Stick-Built House Officially Outdated?

What if building a house could be as simple as building a car? Where a car rolls off the assembly line every few minutes.

BOXABL is inspired by Ford and Tesla’s manufacturing process. Their mission is to make upscale, low-cost housing available globally by bringing assembly line automation to the housing industry. BOXABL’s patented technology allows them to create modular, prefabricated homes that are designed to fold up for transport which significantly decreases the cost of shipping & the complexity of installation.

BOXABL has raised over $170M from over 40,000 investors since 2020. They recently achieved a significant milestone: raising over 50% of their Reg A+ funding limit! BOXABL is now only accepting investment on their website until the Reg A+ is full.*

If you want to join over 40,000 existing investors and be a part of the story, click below to explore investing in BOXABL.

Invest in BOXABL today!

• Genetic-testing firm 23andMe, now in Chapter 11, has notified millions of customers that they must submit a proof-of-claim form by July 14 to seek compensation. The claims cover losses from the 2023 data breach and other grievances, while a proposed $30 million class-action settlement is paused during the bankruptcy process.

• Pope Leo XIV told the College of Cardinals that AI is a new industrial revolution imperiling human dignity, justice, and labor, a concern that inspired him to adopt the name of social-teaching champion Leo XIII. The address builds on recent Vatican documents and Pope Francis’s 2024 warning by urging ethical safeguards so technology serves humanity rather than eroding it.

• A conversation with Jony Ive

• FDA launches agency-wide AI rollout to modernize drug and device reviews

• AI-human task sharing could cut mammography screening costs by up to 30%

• Top AI execs tell Congress the U.S. must boost exports and infrastructure to compete with China

• Yale’s Digital Ethics Center helps states navigate AI policy and regulation

• Mastercard exec says AI agents helping you make your next purchase could be key to solving online shopping’s $750 million fraud problem

• Whoop fitness tracker backpedals on its paid upgrade whoops

• Gamma: Makes beautiful presentations, websites and more

• QAWolf: Cut your QA cycles down from hours to minutes*

• Free event: The AI Readiness Summit*

🤖 Amazon’s Vulcan can feel

Puck-shaped vacuums have given way to machines that see, walk — and now feel. Amazon’s new Vulcan robot adds a tactile sense to the warehouse floor just as startups unleash humanoid androids straight out of sci-fi.

Vulcan combines cameras with force sensors so it “knows” when it touches an object, letting it pick or stow roughly 75% of products without crushing them. It reaches top and bottom shelves, sparing workers awkward ladder climbs and deep bends, and will roll out across the United States and Europe, over the next two years, joining Amazon’s 750,000-robot fleet.

The details:

• Dexterity meets data – Vision + force feedback guides a suction tool that picks from bins of up to 10 items.

• Real-time feedback – End-of-arm “paddles” dial pressure, sliding products into tight slots.

• Human assist – Flags a worker when it hits a tricky item.

• Ergo upgrade – Trials in Spokane and Hamburg cut ladder climbs and deep bends.

Amazon isn’t alone. A wave of humanoid robot startups aim to automate industrial grunt work.

• Tesla Optimus: Gen-3 prototype starts factory trials; Musk eyes mass production under $25 000.

• Agility Digit: Already hauling totes in Amazon pilots; $400 million fresh funding builds a “RoboFab” plant.

• Figure 01/02: Raised $675 million from Microsoft, Nvidia and Bezos; BMW plant trials start this year.

• 1X Neo: OpenAI-backed Norwegian bot aimed at industrial and security work.

• Sanctuary Phoenix: Learns new retail tasks in less than 24 hours; pilots with auto-parts and telecom firms.

Apptronik’s Apollo ($350 million raised) and giants like Toyota and Xiaomi are also in the hunt. Boston Dynamics still sets the agility bar with Atlas, though it remains a demo platform.

Go deeper: Labor shortages bite across the $10-trillion logistics and manufacturing sectors. Venture and corporate investors have sunk at least $1.5 billion into humanoids in the past 14 months. Shared EV batteries, off-the-shelf AI chips and multimodal models are slashing hardware costs and training time, turning once-fragile prototypes into viable pilot fleets.

The hype is thick – dance-floor demos and Muskian superlatives – but the market pull is real. Task-specific bots like Vulcan will dominate the next five years because they fix a clear pain point and slot neatly into existing workflows. Humanoids will follow a classic tech curve: fragile prototypes today, pricey pilot fleets in the coming years, and only later dipping below entry-level human costs. Labor demographics, not sci-fi daydreams, will keep capital flowing. I wouldn’t expect a silver-skinned butler in your kitchen before decade’s end.

🤔 Your thought process:

Selected Image 1 (Left):

• “You shouldn't be able to see faces if they are a silhouette”

• “The fake is well-done, but the poses are too static for the scene.”

Selected Image 2 (Right):

• “The hands on the guy on the right [in the other] image look unformed.”

• “The sand around the players' feet in [the other image] looked fake.”

💭 A poll before you go

Here’s your view on who wins the coding race…

Coding tools:

• “Honestly, option 4: they both rapidly improve and serve different audiences. Design tools will be great for satisfying bespoke creations of "I just need something small..." or "me and my buddies want an app that..." while coding tools will be for the heavy lifting that programmers do now. Just like how Wix didn't completely upset web developers, but gave a new avenue for many who didn't want to pay a full developer.”

Design tools:

• “Coding is the less creative skill out of the two. A tailored visual design that comes with code that works will be better received than smoothly written code that operates a webpage that is outdated”

AI shouldn’t be creating designs or code:

• “There won’t be an advancement on both the design and code side. It will all fall flat when there’s complex logic anyways”

Thanks for reading today’s edition of The Deep View!

We’ll see you in the next one.

P.S. Enjoyed reading? Take The Deep View with you on the go! We’ve got exclusive, in-depth interviews for you on The Deep View: Conversations podcast every Tuesday morning. Subscribe here!

If you want to get in front of an audience of 450,000+ developers, business leaders and tech enthusiasts, get in touch with us here.

_{*Represents a sponsored placement.}

_{*Boxabl Disclosure: This is a paid advertisement for BOXABL’s Regulation A offering. Please read the offering circular here. This is a message from BOXABL}

_{*Reserving a Nasdaq ticker does not guarantee a future listing on Nasdaq or indicate that BOXABL meets any of Nasdaq's listing criteria to do so.}