- il y a 2 jours
Most Recent AI Venture Trends in Silicon Valley
Catégorie
🤖
TechnologieTranscription
00:00Bonjour à tous, mon nom est Adia Strabolev.
00:04Et 10 ans, j'ai commencé le venture capital à l'initiative de Stanford University
00:10en France, en California.
00:12J'ai étudié chaque venture capital-backed startup dans les États-Unis,
00:20chaque angel investisseur, chaque investisseur dans les États-Unis.
00:24Et plus recently, nous avons augmenté cela à l'initiative.
00:29Maintenant, vous allez tous les savoir sur les A.I. trends du Silicon Valley.
00:32Mais avant de parler de cela, je voudrais parler de l'histoire.
00:36Parce que l'histoire va nous expliquer comment vite nous allons être transformés.
00:41Quand je parle de policymakers, quand je parle de l'entreprise,
00:45je commence à commencer par les deux photos,
00:48qui sont une de mes préférées.
00:50Ce sont les photos de l'Esta Parade en New York City en 1900.
00:56Et si vous regardez très bien, vous verrez que vous voyez que
00:59vous voyez que vous voyez que vous voyez que vous voyez que vous voyez
01:01que vous voyez très bien, vous voyez exactement une automobile.
01:07Si vous voyez 13 ans, et je vais vous montrer la seconde
01:10black-and-white photograph,
01:12en 13 ans, c'est la même Esta Parade.
01:15Il y a des thousands de cars, et maintenant, il y a seulement 1 horse.
01:19Donc, en 13 ans, les horses sont remplacés complètement par des cars.
01:24En New York City, en 20 ans, ils étaient complètement remplacés par des cars
01:29dans le monde.
01:30Il y a quelque chose qui est vraiment important,
01:33c'est que, si vous êtes un histoire buff,
01:35vous ne connaissez pas un seul horse breeder
01:42dans le year 1900.
01:44Vous ne connaissez pas un seul carriage maker dans le year 1900.
01:47None of them survived.
01:50All of them were replaced by start-ups of that period.
01:53Some of them became huge companies.
01:56And the investors that invested in those companies
02:01were not the investors who invested in the horse transportation industry.
02:07So, what happened back then is happening today just on steroids.
02:13It's happening much, much faster.
02:15So, it took about 20 years for the horses to be replaced by cars.
02:20It's now taking days for some truly popular programs
02:26to reach critical mass.
02:28And even in hardware, it may be taking 2-3 years
02:32to completely replace the previous predecessors.
02:37So, for founders here, you have a real chance
02:40to become the cars of today.
02:43Building product is much, much faster.
02:47When we look at every single start-up in the United States,
02:51we'll see that many successful companies
02:54are becoming successful about 40% faster
03:00than just 5 years ago.
03:02And if you're the AI-driven start-up,
03:06you're becoming faster, if you're successful,
03:09at around 75% rate.
03:12Which means that one of the best indicators
03:16for the investors, whether start-ups are successful,
03:20is how fast your growth rate is.
03:23Start-ups are growing at about 4 times faster
03:25that investors expected even 5 to 7 years ago.
03:31Here's the data on every single American unicorn.
03:36So, we're following every single unicorn
03:38and producing once a quarter the unicorn report
03:43from Stanford University.
03:45Compared just to some years ago,
03:50these days it takes about one year for future unicorns
03:56to raise the first VC funding.
03:58And it takes less than 4 years to become a unicorn.
04:03And it's about half of what it used to be.
04:07There's a huge signal for founders.
04:09You have no time to lose.
04:12This is from the point of view of the funding AI trend,
04:16is the single biggest trend.
04:18No time to lose.
04:20So, I've met a lot of founders today
04:22walking through the VivaTech.
04:24They're all excited.
04:26And some of them will become trillion-dollar companies.
04:29But my research suggests
04:31that those companies that are very early in the game,
04:35those companies that are able to raise
04:37venture capital funding very, very early,
04:39are much more likely to be successful.
04:44Everybody's talking about AI trends.
04:46But one of the most interesting AI trends
04:48is that there are two types of startups
04:50that are now conquering the unicorn landscape
04:53in the U.S. and globally.
04:56About 50% of unicorns are AI-first unicorns,
05:01which meaning that they completely replaced
05:03the business model and the technology
05:06of the companies that came before them.
05:09But about 50% are AI-enabled unicorns.
05:13They are, in fact, not really in the AI space.
05:16They are in other spaces, biotech, defense, robotics.
05:20But they are using AI-enabled stuff.
05:23And this is where most of the growth is happening right now.
05:27So the past 18 months, most of the startups that became unicorns
05:33in Silicon Valley, more broadly in the U.S.,
05:36and more and more so globally, are AI-enabled startups
05:41that are using AI to replace technologies
05:45and to resolve problems that previous companies
05:48or previous startups tried to resolve much faster.
05:50And I'm seeing this particularly in hardware, in robotics, in defense,
05:57and also in healthcare.
05:59And that is, again, another important message, both for the founders here,
06:04which is you really need not to build just an AI startup,
06:07but apply it to a very specialized business model
06:11and specialized industry.
06:13Investors these days expect, especially in B2B startups,
06:18to get your B2B partnerships very, very early on.
06:22And very often, nobody in Silicon Valley is interested in MVP.
06:27Everybody knows what MVP is.
06:28But they're interested in MLP.
06:32And I'm very often surprised that not so many people still know what MLP is.
06:35MLP is meme-lovable product,
06:38which means that something that your B2B partners or B2C customers
06:45are willing to start using right away.
06:48And that is because of the very, very fast growth.
06:50So that is another message both for the founders
06:53and for the investors about the AI trends right now.
06:59For the investors, it's much more important these days
07:02to support startups very, very early on.
07:05That is because in the past, it took about 18 to 24 months
07:10for the whole cycle in each space to go around.
07:15Nowadays, it takes less than three to six months,
07:19meaning that unless you spot an opportunity very, very early on,
07:23in each specific AI space, you are late to the game,
07:27especially if you're an angel investor or an early stage investor.
07:31And what you would like to do is the most successful investors
07:35are those who not just spot opportunities early,
07:38but those that go against the consensus.
07:43So I started every single VC investor in the United States,
07:47and those that go early and go against the prevailing consensus,
07:53they reap the highest benefits by far.
07:56So if you look at the top 100 VCs in the last 12 months,
08:02there are those that are investing in specific AI spaces
08:05that others will invest only in six to 12 months.
08:10And this very shortening time where you can go against the consensus
08:15is really a critical new AI trend.
08:18This also leads to a huge turnover in investors.
08:22So I've been observing the VC industry for many, many years,
08:26and we looked through the history of 50-plus years.
08:28This is the first time in 50 years where there is such a dramatic turnover
08:35in successful top investors.
08:38It had never happened before, which is if you looked at 10 years ago
08:43and compared to the investors from 15 years ago,
08:45the top 100 investors would be virtually the same.
08:48There will be several new entrants.
08:50These days, every year, there's about a 20% turnover,
08:55which means that every year in 100 top investors,
08:59there's going to be 20 new names.
09:00So investors here, you can get there.
09:04For startups, it's particularly critical to find investors
09:07who can identify their opportunities very, very early.
09:11So what matters most for investors?
09:14Well, it's still about the team.
09:16And for the founders here, one of the most interesting pieces of research
09:20is how investors look at your pitch decks.
09:23But research shows that investors go, look at the pitch,
09:27and immediately go to the last slide,
09:28because the last slide typically is the team page.
09:31My first piece of advice, move it to the second page
09:34after the name of the startup.
09:37And make sure that you tell the investors not what you are doing
09:41but why you are the best to execute on whatever your problem is,
09:45whatever your solution is going to be.
09:48If you look at the research about the factors of success of startups,
09:54the results are absolutely striking.
09:58Overall, in IT, in healthcare, etc., in AI especially,
10:04team is by far the single most important factor of success.
10:08Note something really critical.
10:10I'm not mentioning just founders.
10:13So these days, it's not just about a couple of founders
10:16but the team behind them.
10:17In fact, most of the successful VCs in the Valley tell me,
10:21and my research also supports this conjecture,
10:25is that they demand, these investors, demand seeing the team
10:30and looking how the founders can attract the best,
10:35the most efficient and effective teams that can resolve the problems.
10:40And why that is the case?
10:42Because investors know that those teams will now have an opportunity
10:46to become the first just in six months.
10:49Again, in the past, it took three years.
10:50Now it takes six months.
10:52So the teams you have to form as startup founders very, very early on.
10:57And the teams are behind most failure.
11:01The most important factor in startup failure is also the teams.
11:05And so investors are trying to identify so many reasons to invest in specific teams.
11:13What are the most important reasons?
11:16I mentioned the unicorn report where we also identify every single founder of unicorns
11:21and compare them to founders of non-unicorns.
11:25And we identified about 120 variables that predict the eventual success.
11:31As you can guess, my database is quite popular with investors.
11:35But let me share with you a couple of characteristics.
11:38And even though some of those characteristics might seem to you obvious,
11:44it is the numbers that really matter.
11:46So what is the one characteristic?
11:48It's education.
11:50Now that is the American data, even though we now have European data as well.
11:53And it should not come to you as a surprise that some top American universities are really also behind most
12:03of the unicorns in the United States.
12:07So on the right, in each column is the ranking of the best colleges according to the US News.
12:13But the first column over there is our rank of my team of universities by the number of founders that
12:20graduated from that college.
12:22If you add also the graduate founders, if you add professors and staff, then in fact top 20 universities in
12:32the US will be behind about half of all US unicorns.
12:36I am often asked why there are so many unicorns in the US and so few in Europe.
12:42And one of the biggest reasons is that American universities are structured very differently compared to the European universities.
12:50So education really matters.
12:53And what matters is not just education, but what kind of major you have.
12:58And either two founders or the team, whether there is a major complementing each other or not.
13:04And investors are aware of that.
13:06So for example, for founders here who are trying to search for co-founders, education also does matter.
13:12It is an important factor.
13:14Work experience is another very important predictor of success.
13:20So this is a little bit of technical stuff.
13:24Those are, if you look at all the unicorn founders, and it is becoming particularly important now in the AI.
13:32And again, the reason is because you have to do everything in a much, much faster time.
13:36If you look at unicorn founders, this is their full-time experience before they started a startup.
13:43And then you compare it with founders that did not become founders of unicorns.
13:49The numbers there are what I call the odds ratios, which meaning by how many times you are more likely
13:55to become a unicorn founder.
13:56For example, if you happen to work at Google and then founded a company after you left Google,
14:03you are 2.5 times more likely to become a founder of unicorn.
14:08That is a huge number.
14:11In the world of VC, even increasing your success rate by 3% is huge if you are an investor.
14:17Here we are talking about times, 2.5 times.
14:23And investors do know this.
14:25So I think that having experience, this also extends not just to founders but also to your teams.
14:33And so when you are looking for specific teams, then identifying teams with a specific work experience and specific education
14:42is one of,
14:42well, it's two out of hundreds of variables that were identified.
14:47So let me conclude by saying that the most important trends that I see based on all this research and
14:55all the observations in Silicon Valley is that there is nothing new about AI.
15:00It's just what we observed the trends in the past on steroids.
15:05Everything is going to happen much, much faster.
15:07Let me conclude by saying that based on everything that I've done, I'm going to make a provocative prediction.
15:14It's that 10 years down the road, there will be 10 new $1 trillion companies that right now either don't
15:22exist or relatively small startups.
15:24Maybe here at VivaTech looking to find their first investor.
15:29And this is huge, 10 companies that are going to become 1 trillion plus dollar companies.
15:35And that is because AI is going to accelerate all those trends.
15:39The same many existing companies will disappear or become less successful.
15:44And if you're an investor, you have your chance to invest in one of those 10 companies.
15:50Thank you.
Commentaires