Fresh out of the studio, Benedict Evans, the independent analyst and venture partner from Mosaic Ventures & Entrepreneur First, joined us to talk about his recent annual presentation on the technology business: The New Gatekeepers. Benedict shared his perspectives on the evolution of Web3 and AI given the major shift from the metaverse to generative AI. Last but not least, Benedict delves into the exciting possibilities and potential outcomes for the emerging wave of AI startups and explains what success would really mean for them.
"With both Web3 and VR, if you thought that was interesting a year ago, it's still interesting. Nothing's really changed. Like 98% of all the nonsense has gone away. But if it was early 2022 and you thought VR headsets are going to be a device that two or 3 billion people own, well nothing's changed in the last year to change that. If you think that we might build a billion-scale, consumer software, consumer internet service on a blockchain. Which is really what Web3 means. Yeah, ok. If you thought that a year ago, fine, but you should probably think that is still five years away, but it's just as interesting now as it was then. But it's five years away. Whereas this second wave of AI had been kind of bubbling away very quietly for five years. I mean, the first demo is using generative networks to make faces and things in 2014. And they look terrible. They don't look very interesting. And then suddenly in the last six months, as I've said, suddenly this has worked. Or something's working. And then the growth thing. Well, what is it that's working and what will it become?" - Benedict Evans
- Benedict Evans (@benedictevans , LinkedIn, Personal Blog, Newsletter), Venture Partner of Mosaic Ventures & Entrepreneur First & Independent Analyst
- Since our last conversation, what have you been up to?
- Given that you have been in the business for two decades, asking important questions on key trends in technology from mobile to AI and making investments in the sector, can you share the lessons learnt from your career journey with the younger audience who listen to the show?
The New GateKeepers & Evolution of AI & Web3 with Benedict Evans
- What are the key questions or the narrative which you want the audience to understand in the presentation?
- Who are the old and new gatekeepers that you mention in this year’s presentation?
- Did the pandemic (which is now more or less an endemic) accelerate the demise of the old gatekeepers given that the digital transformation in the past two years was more than the last decade as mentioned by Satya Nadella, the CEO of Microsoft?
- What is the story of Amazon as a new gatekeeper, given that it has three revenue streams: e-commerce, cloud computing and advertising (which has “IPO” in quotation marks recently)?
- How will Amazon be disrupted as a gatekeeper?
- If we recall Marc Andreessen’s famous line, “Software is eating the world”, how did we move from an information arbitrage to software changes the product to now software is a new channel with Shein and Netflix?
- What is the type of opportunities that tech can create now with infinite media (powered by generative AI) and infinite product (which has another business model by targeting the niche or long-tail) as all the old gatekeepers and filters are going away?
- What is the one thing you have learned and now know in building “The New Gatekeepers” that very few people do?
- We experienced a year where web3 and the metaverse went from the darlings of tech to now the doghouse (and of course, it might change given that BTC and ETH are getting popular again). What are the mental models now for those who are still building web3 or crypto and AR and VR?
- In the last part of the presentation entitled “Dreams of the Future”, which will lead towards the current AI wave, what are your perspectives on the unlock of ChatGPT or generative AI as a whole? Is it the iPhone moment, a paradigm shift or the inflated expectations part of the hype cycle?
- What are the questions which we need to ask about this new AI wave?
- Can I build on top of “hallucinations” from generative AI as a feature rather than a bug?
- Given that generative AI is not perfect and has an error rate, how do we control the content which might end up creating disinformation to others?
- What are the type of applications you want to see from this wave of generative AI startups?
- In my recent conversation on why China missed the invention of ChatGPT, one argument coming from those who are watching China and talking to the AI builders in that market is that they just did not go into full research mode with the large language models (LLMs) due to resources constraint, and OpenAI snatched the initiative and successfully launch DALL-E and ChatGPT, now they are being placed in a chasing position. It seems that for China, the key focus is in building LLMs to gain control for both Chinese tech giants and startups similar to OpenAI or Anthropic, but what I see is that in this new wave of AI startups, seems to take the OpenAI, Google or Amazon’s LLM as given and build on top of it, how should we think about AI as layers which we build or not build on?
- Instead of asking you whether AI should be regulated, what are the areas of generative AI that the regulators should think about?
- What does great look like for this wave of generative AI startups?
- Any recommendations which have inspired your life?
- How can my audience find you? Benedict's recommendation: "Why Read the Classics" by Italo Calvino