The week after Silicon Valley Bank’s (SVB) collapse might seem an unlikely time to start talking about a new golden age in innovation and corporate venturing especially when we still have another 18 or so months for this current Age of Ashes to run and further wipe-outs to come.
Already, corporations, such as Google, Stanley Black & Decker and Rolls-Royce, are reducing their internal innovation and venture building units (Area 120, Stanley X and R2 Factory, respectively,) while retaining much of their corporate venture capital (CVC) units where previously established.
As identified in discussions from the first day of the GCVI Summit, SVB had been at the heart of, well, Silicon Valley and hence the dominant venture ecosystem in the world. As Global Corporate Venturing (GCV) wrote in November, the bank probably should have let “a sovereign wealth fund to buy SVB” when its share price first came under pressure as investors recognised the strain the bank was under from its asset/liability duration mismatch.
But, globally, SVB was of more limited importance with a joint venture in China and a separate bank in the UK for much of Europe (now owned by UK-listed bank HSBC after being sold for £1).
SVB had indeed been a core part of the initial golden age between 2011 and end-2016 when it was a competitive advantage for corporations to start or continue using venture tools as so few peers were doing so.
SVB’s support to the GCV events and insights on VC best practices helped enabled the cross-pollination of ideas and networks centred on California and London, then the two main venture hubs, although China was rapidly moving towards parity with the US before the Chinese Communist Party started taking its punch bowl away five or so years ago.
Now, as the central tenets and ideas of corporate venturing expand around the world new, unique forces in generative artificial intelligence (AI) are coming into play to change the dynamic and unleash greater disruption and opportunity.
From innovation stagnation to innovation explosion as AI builds on the central pillars of the scientific methods – trial and error in a measurable and repeatable way – and economic theory – growth comes from an increase in human and physical capital and innovation, however broadly defined between mergers and acquisitions, research and development, and corporate venturing, including CVC and venture building.
In my welcome address, I told the GCVI Summit that corporations can allocate money efficiently between these innovation tools. An increasing amount of money is going into corporate venturing, scaling up the industry. But for there to be a a further order of magnitude jump, we need a change in the way corporate venture finds innovation.
Thomas Edison, inventor of the lightbulb, reputedly said he had successfully found 10,000 ways for something to not work. Author Isaac Asimov in a paper for the Rand Corporation said what was needed was “people with a good background in a particular field, but also people capable of making a connection between item one and item two which might not ordinarily seem connected”.
There were not many of those people between the start of the industrial age to even a few decades ago. Now, there are many more. It is amazing how quickly an ecosystem can grow.
In the mid-2010s, Brazil had just a handful of VCs and several hundred entrepreneurs in a country of more than 250m people. Over the last eight years, however, it has launched more than 150 local CVC units, supporting more than 22,000 startups. A partnership between the ApexBrasil government agency and Global Corporate Venturing, helping to convene and train the ecosystem, has played a role in helping create that.
As the foreword to the World of Corporate Venturing noted: “Venture capital spending accounts for around 1% of total corporate innovation budgets. What would it take to get nearer 10%?.”
Generative AI, such as the newly-released GPT-4 from OpenAI, creates unlimited ideas and ways to quickly test them. Anima Anandkumar, senior director of AI research at Nvidia and Bren professor at Caltech, in a dinner after her fireside chat at the GCVI Summit described how generative AI developed by Caltech had been able to develop a catheter that could reduce the risk of infection by several orders of magnitude in a fraction of time humans using trial and error could have done (if at all).
Similarly, in a blog, Linas Beliunas said: “OpenAI’s GPT-4 has been live for less than 24 hours but it already changed the world. Here are the most mindblowing things you can do with GPT-4 right now: napkin sketch to website, drug discovery, smart contract reviews. AI is the future and it’s going to change every industry forever.”
João Ferrão dos Santos, meanwhile, used GPT-4 to create a new company, writing about it in this LinkedIn post. He says: “I asked GPT4 to become a CEO and create the most profitable company you can build with $1,000 and one human hour per day.”
This is what happened to him in the next two hours:
- GPT-4 told me to start an e-commerce store.
- It chose the brand name AIsthetic Apparel.
- We sell t-shirts made from sustainable cotton. Designed by AI. Obviously.
- GPT-4 told me to attract angel investors for upfront capital.
- I am proud to have raised $2,500 from angel investors Martin Andreas Petersen, Frederik van Deurs and Vasco Conceição, in exchange for 25% of the company. It’s a tough market these days, I took what I got.
- GPT-4 drafted a 10 step business plan and suggested I use Shopify and Printful, selling exclusively print-on-demand.
- We now have a live Shopify store connected to a real life company and a bank account.
- GPT-4 told us what we should ask Midjourney to design, and our first AI-to-AI t-shirt design is complete.
AI provides a potentially limitless idea generation on tap and our distributed computing and ecommerce ecosystems allow these to be easily formed as startups. In this new world, what becomes of the role of the CVC and corporations?
Two things. First, they can still help source the best ideas and help them scale up through the value-added services of capital, customers, suppliers, hiring before an exit. The second — ask the AI to find a solution to the second law of thermodynamics, the one that states that entropy in a system will increase over time. (Those that have read Isaac Asimov’s prescient short story The Last Question, featuring a self-adjusting computer called Multivac, will know that this is the last question the Multivac (or AI) has to answer.)