What we’re thinking about
The cold email that actually got Will, Giant Leap’s Managing Partner, to stop scrolling wasn't polished. His piece on why AI is making founders worse at fundraising ran in Startup Daily this week. Below is an extended edition.
Why I'm craving real human thought over perfect AI prose
By Will Richardson
Lame cold intro from the US.
This was the subject line of a cold email pitch that actually caught my attention. And the content inside followed the same informal, no-polish, subtly self-deprecating tone.
So why did this email, one of the dozens I receive weekly, make me open it, read it, read it again, and reply right away?
As a Partner at Giant Leap, like many other investors, I’ve found the past 12-18 months a mountain of copycat emails and an unavoidable AI-powered pitch-fest from startups. There’s a funny irony that some of AI’s most vocal proponents, startup investors, are now at the bleeding edge of homogeneity being bred by its rampant use.
We’ve received high volumes of inbound for years, but since ChatGPT and Claude became ubiquitous we’re getting orders of magnitude more, and we can usually tell before we’ve even opened it. The subject lines often have tells, like introducing scarcity with “last opportunity”, and then inside it’s very obvious it’s a blanket email blast, with nothing to show they know anything about us. We’ve learned to filter it out almost instantly.
The cost to send an email has now fallen to zero. What’s changed is the friction, not its purpose or value. The email is still the audition.
This sender’s audition had everything I’d hoped for, an outline of real-world operating experience, they had decades, referenced specific companies, shared a personal reason for contacting us, and highlighted a key insight from our website that resonated with them. Although not perfect it was incredibly refreshing and clear it was only written for me.
The misconception by founders is that fundraising is only about volume. The game played becomes “law of averages”, and this is what the dozens of people cold-pitching investors each week are overweighting in the attempt to build their top of funnel.
If fundraising were viewed through a sales lens, it would be viewed as an enterprise sale.
When a founder is looking to undertake a six-figure plus sales opportunities they recognise that the first interaction sets a tone and underweighting the value of first impressions can lead to a perceived lack of credibility. The result is often poorly written email communication that fails to get any real attention when then misses the opportunity to convert to an online or in person conversation where a founder gets to pitch themselves and their product.
One byproduct of an inauthentic and generic cold email is that the reader can conclude that if you can’t succinctly articulate your unique insight and show an ability to execute in an email, it’s hard to believe that you’ll convince investors, recruit employees, win customers, or lead a team.
And it goes deeper than style. It’s not just that everyone sounds the same, it’s that the critical thinking has been outsourced too. You end up with prose that’s perfectly structured around an idea that was never really interrogated. Pre-AI, we noticed orange flags around spelling and grammatical errors, but ironically, post-AI we’re more interested in someone thinking differently and around corners, to show they see a different future they could create and pull into the present. The writing doesn’t need to be perfect, but the ideas need to be sound, and that’s what we’re actually looking for in founders.
I’m not alone in feeling this AI copy malaise. We’ve noticed a distinct movement on the internet whereby people are calling out that this was written by them, not AI, and even adding deliberate mistakes into their communications to appear more human. Even that’s too hard for some: There’s an AI tool that will add typos for you.
LinkedIn, owned by Microsoft, one of the world’s most vocal proponents of AI, has said enough is enough. They are starting to crack down on AI-generated posts, and it’s about time. When people can’t even muster a sincere congratulations on a new role, or award win, and need to use AI-generated copy to muster one up for them, surely something is deeply wrong.
It has lost a lot of its value as a place to find unique ideas, because the cost of producing a post has dropped to zero. When that happens you just get a volume of low-quality writing, and it ends up being noise. We’ve gone from a platform of self-congratulation, people saying “well-deserved” to each other for every award and nomination, to one where everyone’s a thought leader producing the same recycled takes. Honestly, I’m craving something different, and I’m finding it elsewhere. Substack, long-form writing, places where you can tell someone has actually participated in the drafting process. That’s where the interesting thinking is happening right now.
So, where does AI fit into fundraising?
AI is fantastic for refining pitch decks, preparing for meetings, and managing complex cap tables.
We expect that, it’s table stakes, and we expect every company pitching us to have AI embedded in how they build their business too. But we’re not just backing ideas or technology. We’re backing people. Founders who can think for themselves, build exceptional teams, and have a point of view that no model generated for them. It’s the combination we’re looking for, and you can usually tell pretty quickly which side of that line someone is on.
Don’t get me wrong, I use AI every day. But I don’t take its perspectives or summaries as gospel, and I think that’s a distinction worth making. The datasets are easier to manipulate than most people realise. A BBC technology journalist recently spent 20 minutes writing a post on his own website claiming he was the world’s best hot dog-eating tech journalist, based on a fictional 2026 South Dakota championship. ChatGPT and Google picked it up and repeated it as fact. That’s not an edge case, it’s a fundamental limitation of how these systems work. So the question isn’t whether to use AI, it’s how. We use it to challenge our thinking, to apply Edward de Bono’s black hat rather than just get reassurance that our ideas are good. AI is a very capable cheerleader if you let it be, but we’re not looking for sycophantic reassurance, we want to be challenged.
I like to think LLMs are like the greatest cover band in history. They can play anything, flawlessly, on demand. But every song belongs to someone else, and the room knows so, so ‘OK’ or ‘meh’ reactions are fine.
But, a great company, on a great mission, should land as ‘wow’, not ‘meh’.
Like originals, they are created with experiences, scars and the contrarian insights that are experienced by the individual not the AI. And, these experiences and the retelling of them are actually what makes something worth reading.
I know we’ll see new models emerge, ones with access to unique datasets built around real thought leaders, and tools that genuinely understand how you think and help you refine it, but outsourcing your thinking cannot produce value that the world wants.
A model can learn your tone and mirror your cadence, but the thinking still has to start somewhere human.
So before you push send. Pause and think. Is this message actually communicating well on my behalf or, is it just really generic?
Because I’d prefer an inbox with hundreds of irreverent, but sincere subject lines over mountains of messages that just all sound the same and get immediately discarded.
For the road
🗣️The problem with AI slop isn’t the AI. When a junior employee hands in AI slop, Noah Brier argues the failure isn’t theirs. It’s a system that taught people to write by template. AI is a mirror, and right now it’s telling us we’re corporate and boring… but former OpenAI safety researcher Steven Adler has a partial answer: competitive debate. Thinking on your feet, convincing a real human who pushes back. In doing so you learn to think deeper, get more creative and genuinely understand another person’s perspective.
👤 90% of US employers use AI to screen job applicants and most are using tools from the same small pool of vendors. A Stanford study of 3.4m applications found the result: racial bias at scale, and a systemic rejection loop where one rejection makes the next more likely. The problem: these tools are pervasively adopted, highly consequential and opaque to the public.
🌡️ “It cannot be as bad as we thought, but it cannot be as good as we hoped.” Scientists have retired both the best and worst case climate scenarios. Temperatures rising 4.5°C is off the table, but so is any path that stays below 1.5°C. The new best case peaks at 1.7°C for up to 70 years before coming down, and only comes down if carbon removal technology scales in time.
⚡ The world sold one electric car for every four last year. Global EV sales exceed 20m in 2025 and with an oil price shock rewriting the economics, the IEA thinks 2026 could push that to 30%. Australia is moving (13.1%, up from 9.5% in a year) but Europe, China and parts of Southeast Asia are on a different trajectory altogether. Some markets are eyeing 60% by 2035.
🧠 Most AI benchmarks measure model intelligence. None measure what AI does to ours. MIT’s Media Lab convened 80 experts together to start asking 1) whether AI helps people think critically, 2) whether it nurtures intrinsic motivation and skill development and 3) whether it supports genuine human connection. Read the report here.
For your ears
🎧 Orrin built Indi, one of Giant Leap’s portfolio companies, after his daughter was diagnosed with global developmental delay and a year of specialist appointments went nowhere. If you want to hear how it started, in his own words, listen here.
🎧 If you’ve been meaning to get across the AI risk conversation, this is a good place to start. Tristian Harris makes the case that the US and China are not actually competing with each other; they’re both losing to the same thing. Listen here.
🎧 Joey Man from Tenmile, a $250m health and life sciences fund, has been doing deep work on endometriosis and PCOS (now PMOS). A biological challenge and a market failure, and she covers both. Listen here.
🎧 In year 11 of a 13-year drug programme, the trustees of Novo Nordisk blocked a merger that would have cancelled it. That programme was GLP-1, and they created $600 billion by saying no. Eric Ries, author of The Lean Startup, joins Y Combinator to talk about his new book on how mission-driven companies get taken over and the structures that prevent it. Listen here.
Helpful resources
📖 Sean Grealy’s What VCs Actually Want to See in Founders’ Sales Processes interviews five investors, including Giant Leap’s Will Richardson, on where founder-led sales breaks down and what a good handover looks like.
🎥 LaunchVic have released a 40-minute roundtable on raising capital with four Victorian VCs: Flying Fox, OneVentures, Mandalay and Skalata. It covers what investors screen for at each stage, how AI-era traction is read differently and why the best founders start building investor relationships long before they need a cheque.
📊 ANDHealth’s FY2026 Industry Sentiment Survey maps the funding gap in Australia’s digital health sector. A $45b industry where 86% of SMEs cite access to capital as a top challenge, up 26% since 2023.
Giant leaps
🤝 Perx Health has announced two strategic transactions. The Australian business has merged with Navigator Group while the US business has been acquired by Clutch, where it will operate as Clutch Health.
🩺 More Good Days has been accepted into ANDHealth+, Australia’s leading digital and connected health accelerator, one of five companies selected from more than 90 applications to share in up to $9 million in non-dilutive funding across two stages.
⚡ Amber has secured $13.6 million from ARENA to expand its vehicle-to-grid (V2G) program from 50 to 1,000 homes, allowing EV owners to sell stored energy back to the grid. Amber plans to launch a commercial V2G product later this year.
New paths
🩺 Clean Slate Clinic is hiring a Founding Account Executive (Remote).
🤖 Indi is hiring a Design Engineer and a Full Stack AI Engineer (Sydney).
🌱 Trace is hiring a Founding SDR (Sydney, hybrid).
🧻 Who Gives a Crap is hiring a Creative Strategist (UK), Director of Growth Marketing (UK), Head of Brand and Growth (US), Performance Creative Lead (US), and Head of Marketing Operations & Planning (Australia) — all remote.
⚡ Amber is hiring across several roles including a Software Engineering Manager (Optimisation Team), Customer Operations Manager (Voice Team), Customer Experience Specialist (Voice Team) and Strategic Account Manager (Europe).
Save the date
📅 9 June: M8’s Angel Curious’s Reverse Angel Pitch Night - Melbourne. Experienced angels pitch themselves to a room of founders and angel-curious operators. Register here.
📅 16 June: Tangible Finance and Climate Salad’s Scaling Hardtech Without Burning Equity - online. This session is on how to build a capital stack strategy for hardtech companies covering venture debt, equipment financing, asset-backed finance, and how to speak the language of lenders, not just VCs. Tailored for seed to growth stage hardtech CEOs and CFOs. Register here.
📅 17 June: La Trobe’s LaunchPad Lift Off Pitch Night - Bundoora. The top 10 finalists from La Trobe’s 12-week LaunchPad program pitch live. Register here.
📅 17 June: Equity Clear’s Show Us the Data Report Launch - Sydney session. Launch of the report explores why Australia needs better visibility into the investment pipeline, with a focus on the gender funding gap. Register here.
📅 18 June: Equity Clear’s Show Us the Data Report Launch - Melbourne session. Register here.
📅 30 June: Startup Network’s Impact Pitch Night - Melbourne. Four early-stage founders pitch solutions to global challenges, with networking across investors, founders and operators. Register here.
Open for applications
📬 June 14: MAP Accelerator 2026. An accelerator for founders including $20,000 equity-free funding, weekly workshops, dedicated workspace at Melbourne Connect, and access to mentors and EIRs. Apply here.
📬 June 14: RMIT FounderHUB. A free four-week after-hours program for aspiring founders who are RMIT students, staff or alumni, with a pathway to up to $22,000 in equity-free seed funding through the LaunchHUB pre-accelerator. Apply here.
📬 June 15: TiE Women 2026. A global competition for women founders with traction. Open to founders across Australia. Apply here.
📬 June 16: 2027 Westpac Social Change Fellowship. Up to 10 fellowships of $50,000 each for Australian leaders working on social impact initiatives, plus a bespoke leadership program and lifelong Westpac Scholars Network membership. Apply here.
📬 June 21: Monash Startup Sprint. A four-sprint venture-building program for Monash students and recent alumni (graduated after 2024). Bootcamp runs 4–5 July. Apply here.
The tell is always in the first paragraph: technically flawless, zero point of view.