wp-plugin-hostgator domain was triggered too early. This is usually an indicator for some code in the plugin or theme running too early. Translations should be loaded at the init action or later. Please see Debugging in WordPress for more information. (This message was added in version 6.7.0.) in /home4/scienrds/scienceandnerds/wp-includes/functions.php on line 6114ol-scrapes domain was triggered too early. This is usually an indicator for some code in the plugin or theme running too early. Translations should be loaded at the init action or later. Please see Debugging in WordPress for more information. (This message was added in version 6.7.0.) in /home4/scienrds/scienceandnerds/wp-includes/functions.php on line 6114Source:https:\/\/techcrunch.com\/2023\/05\/15\/renowned-investor-elad-gil-on-how-the-great-ai-race-will-likely-shake-out\/<\/a><\/br> Elad Gil, a successful founder and prolific investor, has already been called Silicon Valley\u2019s biggest solo venture capitalist<\/a>, given the massive amounts<\/a> of capital<\/a> he has been investing in recent years, including on behalf of institutions that reportedly include Harvard\u2019s endowment.<\/p>\n His track record goes a long way in explaining his quiet rise. For example, Gil invested in the Series A round of the highly valued payment software company Stripe 11 years ago and has invested in many of its subsequent rounds. He also snapped up stakes in the note-taking app Notion, the cloud collaboration platform Airtable, the military tech contractor Anduril and the design tool Figma, which agreed to sell to Adobe for a whopping $20 billion last September \u2014 though Adobe is still working to sell<\/a> Justice Department authorities on the deal\u2019s merits.<\/p>\n In conversation late last week, Gil \u2014 who occasionally blogs<\/a> but maintains a bare-bones site<\/a> \u2014 declined to answer specific questions about how much he\u2019s managing or some of the amounts he has invested in companies. But the quant VC outfit TRAC<\/a> calls him a \u201csuperforecaster\u201d who has funded at least 155 companies, and whose \u201cbatting average\u201d is .671, meaning 67% of his early-stage investments have raised at least one follow-on round, per TRAC data. (It says at least 30 startups in Gil\u2019s portfolio have become \u201cunicorn\u201d companies, though as Gil himself notes, many valuations are poised to change over the next 18 months or so. \u201cThe really hard times are coming,\u201d he says.)<\/p>\n When we spoke with Gil, we asked about what founders should do if things go from bad to worse. We also talked about his ongoing fascination with AI and some of the early checks he wrote to startups that are now raising serious venture dollars, including Character. AI<\/a>, backed this year by Andreessen Horowitz, Perplexity.AI<\/a>, backed by NEA and Harvey<\/a>, backed by Sequoia Capital.<\/p>\n Not last, Gil shared how he\u2019s using AI to scale up his own work. You can listen<\/a> to our full interview; in the meantime, excerpts of that chat follow, edited for length.<\/p>\n TC: Years ago, you wrote a book titled the \u201cHigh Growth Handbook<\/a>,\u201d about how to scale startups from 10 to 10,000 people. Do you now think there was too much focus on growing so quickly?<\/strong><\/p>\n EG: The focus of my book was around if you hit that magical moment of product market fit, what you do next. . . I think this mantra of growth for growth\u2019s sake really came in mainly during the COVID period. As capital became really cheap and available, people started scaling when they didn\u2019t quite have actual product market fit. They started scaling before they had a lot of customers, or before it was clear that they had a moat that would create some sort of defensibility for their business. I think where things went off the rails was people started raising money many years ahead of where they were. And then they started hiring against that money that they raised instead of hiring against the business that they had.<\/p>\n We\u2019re hearing lots of stories from employees eager to talk about mismanagement inside their companies as things go south. Any advice you can offer to companies about how to scale back without completely blowing up in the process?<\/strong><\/p>\n A lot of the assessment that people are doing actively right now is to ask: Where do I think this is going to be in one or three or five years? And if it\u2019s not working, what should I do? Those are really tough choices to face. People have to make decisions between whether they shrink the team and maybe change direction or do they try to sell the company because it\u2019s clearly not going to work. Do they shut down and return money? If you look at when people raised a lot of the money in the last couple of years, it largely happened in 2021. And if people raise three to four years of money and fundraise when [they] have nine months left, that means a lot of people are going to have to start fundraising at the end of this year. So I feel like the really hard times are coming. I think this is still kind of a little bit of a warm up, or anticipatory period.<\/p>\n In terms of your own investments, can you speak to how much you\u2019ve raised in recent years, and how many companies are now in your portfolio? You were raising $620 million per an SEC filing in 2021 . . .<\/strong><\/p>\n I haven\u2019t really talked too much about it [and] I actually don\u2019t know the exact number [of portfolio companies] right now. Traditionally, I would just invest my own money. Then things started to get bigger in terms of the allocations I could invest behind, so in some cases, I did what are known as SPVs, or single purpose vehicles or investments.<\/p>\n Right now, my model is a bit of a hybrid, where anything that\u2019s kind of small, or if people just want me in as an angel, I can do it personally. If something gets bigger, I can use a fund. If something gets really big, I can use a mix of personal money, fund money and SPVs. I\u2019ve tried to keep a flexible approach so that as I work with different companies at different stages, I can tailor what I do to what they actually want and need. I want to avoid the situation where I have a massive fund and feel the need to go and invest a bunch of money and push it on people and kind of start acting badly.<\/p>\n You were paying attention to generative AI before some others. Have you been taken at all aback by what has been released into the world [on the generative AI front] over the last six to 12 months?<\/strong><\/p>\n For me, the big moment in some sense was seeing things like some really early generative art based on GANs [which is a class of AI and machine learning algorithms]; it was just striking what non-artists could do. Then a little bit after that, when GPT-2 and then GPT-3 came out, that was clearly a moment where there\u2019s such a big step up between them that it was clearly a sea change.<\/p>\n Are you an investor in OpenAI?<\/strong><\/p>\n I\u2019m not involved with most of the things that are happening at the foundation level, but I don\u2019t want to talk about any specific companies or anything.<\/p>\n You sat on a panel<\/a> in LA earlier this month with Ashton Kutcher, whose Sound Ventures just raised a growth fund to expressly back just six or so foundational model companies<\/a> \u2014 three of which it has already invested in: OpenAI, Anthropic and Stability.AI. What do you make of that strategy?<\/strong><\/p>\n There are a handful of companies that are really far ahead in terms of developing these foundation models. And I do think that there\u2019s going to be some scale and capital effects to them, at least for the very most cutting-edge models. So for example, you know, GPT-4 still feels to be quite a bit ahead of everybody else, and obviously, Google has the capabilities to build something against. Anthropic has been iterating on its cloud model. There are a few other players. There\u2019s Cohere and A121 [Labs] and the like. But at least for now, it seems like proprietary models are a generation or two ahead of open source, and if you assume that each model is going to get quite a bit more expensive than the prior generation of model, then you can assume that this trend may exist for at least a couple more years.<\/p>\n That means two things. One is that when there\u2019s GPT-7 or whatever, maybe open source is the equivalent of GPT-6 or GPT-5.5. And GPT-6 is probably going to be incredibly performative. It\u2019s probably going to do all sorts of amazing things. So that leads to the question of what are the really cutting-edge things that you\u2019ll need the most advanced models for, and that\u2019s where I think there\u2019ll be a big chunk of value in the industry \u2014 but I think a lot of it will also just go to the things that are a generation or two behind. And that\u2019s where I think open source will also play a role.<\/p>\n So I kind of view it as a world where we\u2019ll have a handful of very large, closed, proprietary models and an oligopoly market that\u2019s sort of like the cloud world where we have Azure, AWS and Google Cloud as the three big players. I think the models may naturally converge there as well. But then we\u2019ll have a bunch of open source that people will use for all sorts of things in parallel.<\/p>\n Again, for much more with Gil, including why he thinks more founders should consider shutting down and returning capital while they still can, you can listen to our longer conversation here<\/a>.<\/em><\/p>\n<\/p><\/div>\n <\/br><\/br><\/br><\/p>\n
\nRenowned investor Elad Gil on how the great AI race will likely shake out<\/br>
\n2023-05-16 22:27:25<\/br><\/p>\n