How VCs are using AI to find new deals
Welcome to the 10th issue of the Tech Wrap-Up Europe newsletter.
Every fortnight, we'll be sharing interesting stories and expert insights from LinkedIn members.?
In this issue, Sam Shead looks at how VCs are harnessing AI in their day jobs, why some are only backing immigrant founders, which AI startups investors are going wild for and what has happened to Meta's Threads app since it launched in over 100 countries around the world.
It's all very well investing in AI but are venture capitalists practising what they preach and actually using the technology in their day jobs? The answer seems to be yes.?
Some are using AI tools for fairly mundane tasks, like organising busy diaries and responding to founders who want their money. A few, however, are using it to help them find startups to invest in.?
Myrto Lalacos , who works in a VC accelerator, wrote on LinkedIn: "VCs are building proprietary AI algorithms to better process the hundreds or thousands of inbound deals they get per month."
Mattias Ljungman , the cofounder and CEO of London-based venture capital firm Moonfire , is one of them. "We employ AI to analyse massive data sets and identify potential investment opportunities," Ljungman told LinkedIn News.?
When hunting out potential deals, Moonfire's AI models aim to take into account factors like market size, product innovation and team capabilities, Ljungman said.?
"These models evaluate the vast field of startups and provide a curated list of promising companies that align with our philosophy , allowing us to focus our resources effectively," he said. "This AI-assisted approach has revolutionised how we find and evaluate potential investments."
Lungjman, who previously helped Skype co-founder Niklas Zennstr?m set up VC heavyweight Atomico , claims that the AI approach has been "quite successful" so far, adding that it has enabled his team to sift through a much larger pool of potential investments.?
"This has led to a much more inclusive approach to VC, allowing us to spot potential wherever it may be," he said. "As a result, we’ve invested in close to 50 companies where we have had follow-ons from and co-invested with Sequoia Capital , Accel , Index Ventures and General Catalyst ."
But things don't always go to plan. "Just like anything else, our AI isn't perfect," Ljungman admits, noting that sometimes the AI will recommend startups that aren't a good fit, or miss ones that are.?
"These bumps in the road do happen, but we take them as golden opportunities to learn and grow," said Ljungman. "When we miss the mark, we dive deep into why it happened, and we make sure our AI learns from it too. This way, every stumble is really a step forward in making our AI smarter. And honestly, despite these hiccups, the pros of using AI massively outweigh the cons."
VCs don't need to worry about losing their jobs just yet though. "While AI is an exceptional tool, it cannot replace the human ability to connect with and understand a founder's passion and vision," Ljungman said, pointing out that an AI can't participate in brainstorming sessions that shape a company's early days or sympathise with the struggles faced by startup founders.?
Ljungman puts it like this: "We're merging the analytical power and efficiency of AI with the intuition, empathy, and experiential wisdom of our human team."?
Moonfire aims to put the "first money" into early-stage startups that then go on to raise more capital at higher valuations. Notable successes from the firm's first fund include HR startup Humaans and business intelligence software provider Lightdash , which have both gone on to raise multi million-dollar series A rounds recently.
Elsewhere, other VC firms that are using AI to find deals include the likes of InReach Ventures in London, as well as SignalFire and Goodwater in the US.?
There's also Hoxton Ventures in London, which has made a name for itself (and money for its LPs) by backing the likes of Deliveroo and Darktrace who are both now listed on the London Stock Exchange .?
Hoxton's AI engine is called Robscore and the company invests a "ton of data" into it, according to partner Hussein Kanji .?
"But [the] first rule of AI is not to talk about your AI," he added.?
Mari Luukkainen noted on LinkedIn that AI can actually help to increase diversity. "AI doesn't care if a founder went to Harvard or if their startup's located in a tech hub," she said. "It's not swayed by a founder's charisma or the latest trends. It just looks at the cold, hard data.
However, not everyone is convinced that AI has a serious role to play in venture investing just yet.?
Shmuel Chafets , a general partner at Target Global , told LinkedIn News that his firm isn't putting too much faith in AI, adding that some funds who say they've fully embraced it may be overstating the reality.?
"Every time some technology like this comes…data mining… the last round of AI stuff…there are funds that come up and say we're doing this," he said.
"We will use it, for sure. But…the core investment decision won't be there," Chafets said.?
On August 4, 2020, a large amount of ammonium nitrate stored at the Port of Beirut in the capital city of Lebanon exploded, killing over 200 people. The tragic event got tech investor Ramzi Rafih thinking about what he really wants to do in life.?
After starting out as an investment banker at J.P. 摩根 in 2007, Rafih then moved into private equity, taking on roles at Silver Lake and then KKR , where he spent over seven years. But he wanted to back companies at an earlier stage. He'd been turning his eye to angel investing around the same time that the disaster struck in Beirut, spending his weekends flicking through startup pitch decks and investing in a couple too.?
"I quickly realised this was very interesting," he said. "You would get close to founders very quickly. You can help them in a meaningful way by making one introduction. There's something about the early stage – where you believe in people before they become successful – that has a special type of reward."?
Rafih has just raised $10m for his own early stage VC fund : No Label Ventures . Unlike most other VC firms in Europe, the firm is only backing non-European immigrant founders. Rafih is originally from Lebanon. Born to immigrant parents during the Lebanese civil war, he then immigrated to France for his studies and later to the UK for his career.?
He says his personal story is one of the reasons he decided to back immigrant founders but he's also incredibly bullish on their potential.?
"Regardless of whether you start a company or not, being an immigrant is like being an entrepreneur," says Rafih. "You leave your country, you set up shop in a new country where you very often don't know anyone, you sometimes don't speak the language…going through that journey is hugely formative and I think it prepares you very well for becoming an entrepreneur later on."?
In the US, 55% of unicorns ($1bn companies) have been founded by immigrant entrepreneurs and Europe is poised to follow suit, according to Rafih.? "We're nowhere near Silicon Valley," he? said before going on to note that there's been a lot of progress in Europe over the last 10 years.
Indeed, there have been some big exits, and some of that money is now being reinvested. There are also multiple examples of role model entrepreneurs who have started with nothing and built amazing companies, Rafih said.
So far No Label Ventures has backed around seven startups with cheques of around $150,000 to $250,000. The founders of the startups hail from the likes of China, India and Morocco.?
领英推荐
Another VC firm with a similar premise is Blue Lake VC , which is investing in immigrant founders in the UK that are focused on business-to-business (B2B) and software startups.?
"Immigrant focused VCs are well established in the US," David Gilgur , founding partner at Blue Lake, told LinkedIn News. "In the UK we are lagging."
"Our reasons (for focusing on immigrant founders) are pragmatic," Gilgur said. "There is data and research that suggests that immigrants outperform local founders."
"In the industry that is dominated by cliques and lack of diversity, we get to do some good," he added, noting that his firm runs an "International Founders Office Hours" in an effort to breach the gap between UK and early stage international startups.
Investment into generative AI companies that build tools capable of creating text, images, videos and other content is continuing to soar, with billions flowing into startups focused on the buzzy new technology in the past few months.
On June 29, 微软 (LinkedIn's parent company), 英伟达 , Bill Gates , and Eric Schmidt backed Google DeepMind co-founder Mustafa Suleyman and LinkedIn co-founder Reid Hoffman 's new AI startup with $1.3bn. It is now valued at $4bn, despite being just one-year-old.?
The company, known as Inflection AI , has developed an "empathetic" chatbot called Pi that's designed to be more friendly and personable than other AIs. Users can chat to Pi on the company's website or app, as well as via WhatsApp, SMS, Instagram DMs and Facebook Messenger.?
"You can talk to Pi about anything and it is designed to be curious and patient, and happy to help you in every way it can," Suleyman told LinkedIn News. "It is here to help you process thoughts and feelings and express yourself."
You can ask Pi help how to host a social gathering on one platform and it will check in on how the event went when you catch up later on another. Suleyman thinks that everyone will eventually have a personal AI to support them and act as their personal assistant, or chief of staff.?
"Conversational AI is the new digital interface with the web," Suleyman added. "It is a new kind of clay and there will be new products and services built using it. But there will be millions of AIs in the world for a wide variety of uses, from enterprise to more practical tools."
On the topic of other AIs, Inflection has stiff competition as several other generative AI startups have raised vast sums to help fund their aggressive growth plans. AI firms tend to spend most of their money on staff and servers. The most talented AI researchers can command six and even seven figure packages, while the most advanced AI models are typically trained on copious amounts of computing.?
Elsewhere, OpenAI has raised over $11bn from investors including Microsoft and is thought to be valued at over $27bn, making it the world's most valuable AI startup. The San Francisco-headquartered firm, led by co-founder and CEO Sam Altman, has become a household name in the AI industry following the launch of its ChatGPT product.?
Anthropic , another rival, has raised over $1bn from investors including 谷歌 , Salesforce, Spark Capital , Sound Ventures , Zoom Ventures Ltd and others. Founded in 2020 by former OpenAI researchers, Anthropic touts itself as a firm with a focus on cautious AI development. It has developed an AI assistant called Claude and is thought to be worth around $4.1bn. This week, the startup is releasing a new AI chatbot , Claude 2, in the US and the UK.?
Elsewhere, startups like Hugging Face , Stability AI , Cohere , Adept , Character.AI , Scale AI and Abnormal Security have also raised substantial amounts of money from investors.?
"Not all ‘GenAI’ startups are going to get this level of funding – but for a few, the ‘billion dollar baby’ club – it's all gravy," AI expert Nina Schick told LinkedIn News. "They need these gargantuan sums of money due to fierce competition on hardware and talent."
Asked why investors are betting so big on generative AI startups at the moment, LinkedIn Top Voice Nathan Benaich said: "In any wave of innovation, investors allocate inordinate sums of money to the latest new thing, in the hope that the economics will sort themselves out later."
"The vast majority of the $14 billion VC firms have invested in generative AI has gone to large model companies. While this vindicates the 10-year overnight success of AI entrepreneurs and researchers, the market is probably overheating."
But this isn't necessarily a bad thing or proof that generative AI is a bubble, according to Benaich.?
"Experiments will be run, attrition will take place, and a select number of strong businesses will survive," he said. "The only guaranteed winners in any gold rush are the people selling the shovels. With the consumer internet, it was Google and Facebook . With large model generative AI, it's currently Nvidia. However, the race is still in its early days, and there will be a vast array of non-obvious winners to come."
AI advisor Henry Ajder told LinkedIn News there are several reasons why investors are going wild for generative AI startups.?
"Compared to other recent hyped technologies (for example NFTs or the metaverse) generative AI has already provided utility and driven meaningful adoption," he said.
While there's still hype around generative AI, the gulf between is much smaller than other tech trends we’ve seen in the last five years or so, Ajder said.?
"With the huge variety of actual and potential uses of generative AI (particularly large language models), there is a growing sentiment that this is the early stages of a global paradigm shift akin to the emergence of the internet," he added. "Seeing huge tech companies like Microsoft dramatically redirect their entire strategies to prioritise generative AI, Google issuing a ‘code red’ following its dominance of online search being disrupted, and OpenAI’s record breaking two months to 100 million users only reinforces investors’ view that we are seeing the next generation of tech titans being born. Securing a foothold in that next generation, which 彭博资讯 forecasted will create $1.3tn in revenue by 2032, is clearly a future investors feel they cannot afford to miss out on."
Perhaps unsurprisingly, many of the best funded AI startups are clustered around San Francisco and Silicon Valley, but London has emerged as the main outpost in Europe.?
OpenAI announced that it plans to open an office in the UK capital this month, hot on the heels of Anthropic. They'll be competing for top AI talent with Google DeepMind, which employs around 1,000 people in King's Cross.?
Maria Luciana A. , head of responsible AI at 普华永道 , noted that "it remains to be seen if the enthusiasm will be curbed by the next big law suit or a AI Cambridge Analytica scandal."
Jan Voss – Jan is the managing director of Cape May Wealth. He talks about family office investment strategy and VC trends and news .
Sara Wallin – Sara is the CEO of Chalmers Ventures, a Swedish tech incubator and investment firm. She talks about recent investments and trends in the European startup space.?
Luciana Lixandru – Luciana is a partner at Sequoia Capital in London. She was previously at Accel Partners. She talks about tech talent and where the firm is placing its bets.?
Patrick Ryan – Patrick is the co-founder and COO of VC firm Odin. He talks about angel investing and investment trends.
Get Hired UK – A fortnightly update dedicated to advice and insights on landing your next role and progressing in your career.
Get Hired Europe – A newsletter focused on helping you land your next job.
A bi-weekly digest of news, views and insights from the world of tech written by Sam Shead .
Director at ABC Consortium Limited (Brand:ABCXchange)
9 个月It is interesting that a VC is spending hundreds of thousand dollars to develop a product and incur maintenance expense every year rather than use the same product with post matching advanced scoring system for no development expense and a fraction of the annual maintenance cost. One can learn more about Alfred at abcxchange.com (https://abcxchange.com/alfred)
Innovation Driven | Results Oriented
1 年After reading the articles, I wonder about differentiation. If all VC’s go for some sort of AI support for choosing opportunities in their deal flows, will them all be adressing a smaller pond? A bit like Waze that creates traffic by sending all drivers to the same small shortcut?
Founder at The Developments Group 07492070000
1 年Ai misconception is vey dangerous because it’s been used to give impression that the Ai is everything while it’s just picking up the options from its programmers and no programmer knows or have all options
Project Manager | Focus on Business Growth, Marketing Strategies, and Data Analysis & Visualisation
1 年very interesting!