Hey — It’s Nico.

Welcome to another Failory edition. This issue takes 5 minutes to read.

If you only have one, here are the 5 most important things:

• Linqto, the fintech marketplace that promised private investing for all, filed for bankruptcy — learn why below.

• What people are vibe coding (and actually using).

• Elon Musk’s Grok 4 AI models set new benchmark records.

• Space startup Varda raises $187M in funding to make drugs in orbit.

• Isomorphic Labs, a startup that wants to use AI to “cure all diseases” has begun human trials — learn more below.

This issue is brought to you by dofollow.com — the link and citation building agency that helps your startup rank on Google, ChatGPT, Perplexity, and Gemini.

The Fastest-Growing SaaS Brands All Have This in Common.^AD

Your SaaS won’t win in 2025 by being better—it’ll win by being discovered first.

AI and search tools now decide what your customers see, and they trust brands with backlinks from credible sources.

That’s where dofollow.com comes in.

We help B2B SaaS companies earn high-authority backlinks from niche-relevant sites—so your brand shows up where it matters.

What you'll get:

• A dedicated strategist in your niche

• Manual, relationship-based outreach 

• Link monitoring + free replacements

• No long-term contracts—just results

Trusted by Surfshark, Heymarket, Wiz, and Pitch.

Get backlinks your buyers (and AI) notice.

This Week In Startups

🔗 Resources

What people are vibe coding (and actually using).

Waves, dams, and rivers: How demand works at scale.

You're doing quarterly planning wrong.

The art of cold-emailing a billionaire.

📰 News

Google adds image-to-video generation capability to Veo 3.

YouTube prepares crackdown on ‘mass-produced’ and ‘repetitive’ videos made with AI.

Elon Musk’s Grok 4 AI models set new benchmark records.

iOS 26 beta 3 dials back Liquid Glass.

💸 Fundraising

Space startup Varda raises $187 million in funding to make drugs in orbit.

Vehicle repair management startup ServiceUp raises $55M.

Also raises another $200M to build e-bikes.

Tailor, a ‘headless’ ERP startup, raises $22M.

Fail(St)ory

Shares Not Included

This week, Linqto, the fintech marketplace that promised to open the gates of private-company investing to everyday investors, filed for Chapter 11 bankruptcy.

Not long ago, Fintech Twitter was showing off screenshots of Linqto’s dashboard and bragging about “owning” SpaceX shares for the price of a new laptop. 

Now the app is dark, withdrawals are frozen, and regulators are combing through a maze of special-purpose vehicles to see who actually owns what. It’s a fast fall from hype to courtroom.

What Was Linqto:

Linqto wasn’t a fresh-out-of-YC rocket ship—it actually dates back to 2010. For years the company tinkered with ways to let small investors play in the same sandbox as big VCs. 

The breakthrough came around 2021, when late-stage valuations were going vertical and every group chat wanted pre-IPO shares of SpaceX or OpenAI.

Here’s how the pitch landed on a would-be user:

1. Open the app, snap a selfie, upload a tax form.

2. Browse a carousel of dream startups—SpaceX, Ripple, Neuralink, you name it.

3. Tap “Invest.” Minimum $5 k, no management fees, instant digital receipt.

Behind the scenes, Linqto pooled each deal into a new SPV, wired money to an early employee or secondary broker, and dropped a shiny “You now own X shares” line item on your dashboard. On paper it felt like Robinhood for private equity: fast, cheap, and brag-worthy.

The model caught on. By 2024 Linqto claimed users in more than 100 countries and touted headline assets in the hundreds of millions—roughly half of that sitting in Ripple alone. Marketing leaned hard on FOMO: screenshots of rising valuations, push-notes about “last call before this allocation closes,” and webinars with founders of the featured unicorns.

But the simplicity on the surface masked a messy basement. Each new SPV added legal complexity; investor accreditation checks were thin; and the chain of title from seller to end user wasn’t always airtight. Still, as long as private-market prices kept climbing, no one asked too many questions.

The Numbers:

• 📅 Founded: 2010

• 💰 Capital raised: ~$3 million

• 👥 Community reach: 100+ countries of accredited investors

• 💵 Minimum ticket: $5,000 per deal, zero platform fees

• ⚖️ Bankruptcy filed: July 8 2025, citing “insurmountable operating challenges”

• 🪙 Key asset: An estimated $500 million worth of Ripple equity locked up.

Reasons for Failure: 

• Legal guardrails ignored: Everything Linqto offered was a security, yet investor verification and disclosure rules were glossed over to keep signup friction low. Once the SEC and DOJ opened joint investigations, momentum turned into quicksand.

• A cap-table that didn’t add up: An internal investigation concluded that many customers held indirect contractual claims, not actual shares, in the underlying startups. In practice that meant no voting rights, no transferability, and—most damning—no guarantee their paper would map to real equity if an IPO happened.

• Liquidity Mirage: Half a billion dollars in Ripple stock made for a great headline, but an illiquid holding doesn’t pay bills. When trading was suspended in March, the company’s most valuable asset was effectively frozen, limiting options for a soft landing.

Why It Matters: 

• Compliance isn’t optional; skipping it turns growth into legal quicksand.

• Clear, verifiable ownership keeps investors calm when markets get noisy.

• One bad headline can erase a fintech’s entire trust economy.

Trend

AI wants to cure all diseases

There’s been a lot of noise in the AI and biotech worlds lately, but here’s something worth paying attention to: Alphabet’s Isomorphic Labs just announced it’s staffing up for its first human clinical trials. After years of behind-the-scenes R&D, their AI-designed drugs are finally heading into real-world testing.

That’s a big moment. Isomorphic isn’t just another biotech startup, it’s an offshoot of DeepMind, and it’s armed with AlphaFold, one of the most powerful tools ever created for understanding biology.

So this week, let’s dig into what Isomorphic is actually building, how it works, and what it could mean for the future of medicine.

Why It Matters:

• AI enters the clinic: For the first time, drugs designed by AI systems are about to be tested in humans. This could mark the start of AI's real-world impact on medicine.

• A new biotech playbook: Isomorphic isn’t just optimizing drug discovery—it’s rebuilding the pipeline from scratch, turning drug design into a computational problem. If this works, “AI-first pharma” could become its own category.

• Lower barriers, wider access: By replacing early trial-and-error with in silico modeling, this approach opens the door to leaner, faster biotech R&D.

What is Isomorphic Labs?

On paper, Isomorphic Labs is a London-based biotech company owned by Alphabet. But under the hood, it’s basically DeepMind’s bet on turning its AI breakthroughs into a drug discovery engine.

The company was spun out in 2021, but the real origin story goes back further—to DeepMind’s AlphaFold project.

AlphaFold didn’t just make headlines; it changed biology. It solved a fifty-year-old problem in molecular biology: how to predict the 3D shape of proteins from their amino acid sequences. That’s huge, because the shape of a protein determines how it functions—and more importantly, how it can be targeted by drugs.

The latest version, AlphaFold 3, takes things a step further. It doesn’t just predict protein shapes; it models how those proteins interact with other molecules like DNA, RNA, and potential drug compounds. This makes it a much more powerful tool for designing new medicines—not just understanding biology, but manipulating it.

And that’s where Isomorphic comes in.

Isomorphic’s big idea isn’t just to use AlphaFold as a research assistant—it wants to build an entire drug development pipeline around it. That means combining machine learning scientists with pharma veterans, designing drugs both for pharma partners and in-house, and optimizing every stage of the process from idea to clinical trial.

Why This Is a Big Deal

Let’s zoom out for a second. Drug development is expensive, slow, and mostly unsuccessful. Pharma companies often spend a decade and billions of dollars to bring a single drug to market—and even then, 90% of candidates fail once human testing begins.

That failure rate is a killer. And that’s exactly what Isomorphic thinks it can fix.

By simulating how drug compounds interact with their targets before anyone steps into a lab, the idea is that researchers can weed out bad ideas early and double down on the ones with real potential. You don’t just save money—you increase your odds.

That’s not speculation. This is already happening. What’s new is that we’re about to find out how well it works in real life. According to CEO Demis Hassabis and COO Julian Murdoch, their AI-generated drugs are already being fine-tuned and prepped for the first phase of human clinical trials.

And it’s not just about making better drugs—it’s about making the whole process more scientific, more predictable, and less like gambling.

And here's the bigger ripple: this kind of AI-first pipeline lowers the barrier to entry. You don’t need your own wet lab, your own fleet of PhDs, or a billion-dollar war chest to start exploring drug candidates. With the right models and access to compute, smaller startups can run early simulations, explore novel compounds, and contribute to the drug discovery process in ways that were previously out of reach.

If these trials go well, it’ll validate a new model for biotech: one where drug pipelines are built first as software systems, not slow iterative experiments.

Help Me Improve Failory

That's all of this edition.

Cheers,

Nico