post

Footer About

Footer Contribute

Footer Get Involved

Footer Links

Mind the Product

Customer

How behavioral economics can improve product strategy

Product Roadmap

Why most product roadmaps fail: Phil Hornby (Product Leadership Coach) – ProductTank Porto

B2B product management

Artificial Intelligence (AI) & Machine Learning (ML)

Building Product Teams

Business Strategy

Data Driven Product Management

Digital Transformation

Diversity and Inclusion

Minimum Viable Product

OKRs

Prioritisation

Product Design

Product Research

Product Development Process

Product Ethics

Product Discovery

Product Leadership

Product Management Career

Product Strategy

Stakeholder Management

🌎 <a href="https://www.mindtheproduct.com/world-product-day/" style="color:#FFFFFF;text-decoration:underline;">Happy World Product Day!</a> Be sure to celebrate product, yourself, and your peers today! 🎉

Join us at #mtpcon London in 2025 – mark your calendars

Connecting fans to creators at Spotify (Dariusz Dziuk, Product Lead, Spotify)

Why teams often struggle with OKRs: A conversation with Jeff Gothelf

Deep dive: Crafting your product vision and mission

5 ways to invent and simplify as an AI/ML product manager

Eira is an editor for Mind the Product. She's been a business journalist, editor, and copywriter for longer than she cares to think about.

Eira Hayward

There are artificial intelligence (AI) startups everywhere you look at the moment, and they’re attracting lots of excitement and interest from investors. Perhaps you’ve read about <a href="https://mistral.ai/" target="_blank" rel="noopener">MistralAI</a>, the four-week old French startup formed by former Deep Mind and Meta staffers that just secured $113 million of seed funding to compete against OpenAI?
Despite an investment environment that is widely held as being very tough, with the days of easy access to capital long gone, there’s a rush of capital to AI projects that is pushing their valuations upwards and creating a very competitive market for AI startups.
So much so that Sheila Gulati, managing director of venture capital investors Tola Capital recently told <a href="https://techcrunch.com/2023/07/02/investors-investment-cadence-2023/" target="_blank" rel="noopener">TechCrunch</a>: “We have closed deals in these spaces in 2023, but the frenzy around AI has definitely meant a lot of capital has rushed into this market. The result has been that we have backed off certain deals based on valuation, and we expect this to continue in the AI world. It has meant fewer deals overall.”
Isabelle O’Keeffe, Partner and Head of Origination at <a href="https://surevalleyventures.com/">Sure Valley Ventures</a>, an AI-focused venture capital firm working with early-stage startups in the UK and Europe, says she’s seeing lots of excitement around large language models(LLM) and a sudden explosion of products for business and consumers. “We’re doubling down on our AI focus with our new fund,” she says. “LLM is super hot. There’s been a flight of investors from crypto to AI and an influx of investors without as much expertise. We’re seeing large valuations, particularly in the US. Typically Europe is not as hot, although we have a $100million plus valuation for the month-old French company [MistralAI] working on generative AI.”
Let’s also remember that the vast majority of AI projects don’t succeed. Just look at the numbers. Most startups fail &#8211; for example, the success rate of Y Combinator startups is about 9% or 10%, based on the accelerator’s list of its most successful startups in 2022. Most AI projects fail &#8211; a 2019 report from Gartner estimates the failure rate is 85%.
Lots of hype and excitement, investors piling in, valuations pushing upwards, high failure rates &#8211; it has all the makings of an investment bubble, similar to the<a href="https://www.investopedia.com/terms/t/tech_bubble.asp"> dotcom and crypto bubbles</a> of the past. Is it a bubble that’s likely to burst? And if not, what&#8217;s different this time?
Isabelle puts AI into three categories for investment purposes: the foundational layer, the infrastructure layer and the application layer. The foundational layer is where OpenAI, Google and so on play. “Generally, the bets here have already been made &#8211; it requires large amounts of capital to play here,” she says. Her firm focuses on startups working on the infrastructure layer: “There’s a lot of interesting stuff &#8211; using AI to augment or scale products &#8211; going on here,” she says, and you can see the types of investments Sure Valley Ventures is making <a href="https://surevalleyventures.com/companies/">here</a>. The application layer is where it can get tricky. “A lot of people feel pressure to put AI in their roadmap,” she says.
Like Isabelle, Jacob George, Head of Product at <a href="https://foundersfactory.com/">Founders Factory</a>, is wary of startups building products as “a wrapper around ChatGPT or Bard”. He says: “Essentially, you’re outsourcing your data science, your algorithm… there’s a lack of IP. If you don’t own your IP you’re just a wrapper around an existing model that everyone has access to.”
Jacob says the current interest certainly has the makings of an investment bubble. He says that there’s so much talk of AI and adding algorithms in the pitches he sees, that “it’s a bit of a red flag that we’re creeping into a bubble”. This time however users/ mass consumers can understand and see how AI can be applied. He says: “The main difference I see this time compared to other bubbles is the product user friendliness.”
Richard Abrahams, CIO and Co-founder of investment platform <a href="https://www.viasprout.com/">Sprout</a>, comments that investors have learned the lessons of the past. He says: “There&#8217;s a lot of headlines, but actually, the funding environment is incredibly difficult at the moment. Whether you&#8217;re using AI or not, the expectation of the majority of investors is still ‘show me revenue, show me strong unit economics, I want to see that you&#8217;re not burning money left, right and centre’.” Jacob adds that the days of VCs writing cheques with no real sight of profitability are long gone: “You really need to have a path to profitability, which was never used to be the case in early stage ventures.” While a startup entering Founders Factory’s accelerator program wouldn’t be expected to show a path to profitability, Jacob says he would want to see a product mindset and that the right things were being tested.
A lot of investors are looking at AI from the other side of the coin, says Richard, and asking about the disruption that it could cause to businesses in their portfolios. He says: “You have to look at what people still do that&#8217;s very manual and what can be effectively replaced or supported by automated features. And that&#8217;s what AI will do really well.”
AI investment funding is having a frothy phase and lots of hopeful startups won’t make it. But perhaps it won’t become a bubble that bursts, because, as Isabelle says, “AI is a way to build products faster, cheaper and easier”.
So what can you do to improve the chances that your startup succeeds? What are VCs looking for from an AI startup?
Do you have intellectual property or proprietary data? A great start is a proprietary data set around a target customer or vertical market. Says Isabelle: “For example there are businesses in the mental health space that have been research-led and are now commercialising their data and productising it in an efficient manner using LLM.”
Is there potential for mass adoption? Says Jacob: “From a product perspective I look for something that anyone with no coding experience can start to use. How can the general public access it?”
Will people buy it? Says Richard: “I was talking to a founder of a software company last week that listed for $4.5 billion. While the company disrupted bigger players, the founder said they weren&#8217;t better, but they listened more. They spoke to and listened to their customers a lot more than the bigger players.” Investors are only interested because they want to make a return, he says, so it always boils down to whether people will buy what you’re selling and how they will buy what you sell.
<h3>Further reading</h3>
<ul>
<li><a href="https://www.investors.com/news/technology/artificial-intelligence-stocks/">AI Stocks: Tech Giants, Cloud Titans, Chipmakers Battle For An Edge</a></li>
<li><a href="https://www.beauhurst.com/blog/ai-startup-companies/">The Artificial Intelligence Top 50</a></li>
</ul>

core/freeform

How Google makes AI work at Enterprise scale – Miku Jha (Director, AI/ML and Generative AI, Google Cloud)

LLM workflows for product managers: 3 key takeaways (Niloufar Salehi, Assistant Professor at UC Berkeley) – ProductTank SF

The future of product management: Insights from ProductTank San Francisco

Product lessons learned making early moves with AI in media: Lindsey Jayne (CPO, Financial Times)

A year with ChatGPT and product innovation: Navigating the AI landscape

How to keep your head about generative AI (when everyone is losing theirs) by Claire Woodcock

AI startups: a frothy market or a bubble that could burst?

Eira Hayward

Read more

Comments