CGTN Europe interviewed Rob Kniaz, founding Partner of H-Tree Capital
Category
🗞
NewsTranscript
00:00Hello, welcome back.
00:01Chinese tech giant Baidu has reported a surprise jump in revenue
00:05despite intense competition in artificial intelligence
00:07and an ongoing economic slowdown.
00:10First quarter revenue rose 3% to $4.5 billion.
00:15Net income for the first three months of the year
00:17came in at just over $1 billion,
00:19which was up 42% from the same period last year.
00:23And as part of that AI competition,
00:25Baidu has axed subscriptions for its chatbot Ernie,
00:28making it free for individual users.
00:31Rob Nyes is founding partner of H-Tree Capital,
00:34a London-based technology venture capital firm.
00:37Everyone's riding this wave of AI right now in the frenzy
00:39to apply it to different areas of opportunity.
00:42Baidu, again, is fairly unique, like Google or Alphabet,
00:45in a way that they also have self-driving cars,
00:47so they can also apply AI across their various business units.
00:51So is that going to be where the company's growth is,
00:53do you think, the AI division?
00:56I think more likely than not.
00:57I mean, obviously, they have lots of different bets on,
00:59like Alphabet does,
01:00but I think, you know, in the clear and present sense,
01:02the most obvious opportunity is going to be all around AI
01:04and spinning that as many different ways as they can.
01:07When it comes to their chat GPT equivalent,
01:10we'll call it Ernie,
01:11they've made that free to individual users.
01:14That sounds like a counterintuitive move.
01:18I think it's a way to get awareness and usage up,
01:20and frankly, to get them better data
01:22that they can start tuning themselves.
01:23I mean, when you're building an AI model,
01:26there's training data and then there's real user data,
01:28and frankly, for someone starting from behind,
01:30like they are with Ernie,
01:31nothing really benefits you better
01:32than having real actual user interface.
01:35So I think it's a smart move for them to build Mindshare
01:38and to show that, frankly,
01:39they can still come up and build a winning product.
01:42It's obviously a very competitive space, AI,
01:44particularly in China.
01:45How competitive is Baidu in that arena, do you think?
01:48I think that they're a player.
01:51I mean, they're one of several
01:52that are all leapfrogging each other.
01:54Nearly week by week,
01:56one will do another launch that upends the other.
01:58So I think it's still an open stakes race
02:00that it's unclear who will be an ultimate winner
02:02or whether there'll be several
02:04around the same level of quality.
02:06But I think Baidu has just as good a shot as anyone does,
02:08similar to how the US players are shaking out now.
02:11And you mentioned Baidu's autonomous ride-hailing service,
02:15Apollo Go.
02:16Where does Baidu stand in that market?
02:20Yeah, they're coming from behind.
02:22They're looking at entering Europe now.
02:24I think it's interesting and it's an ambition
02:26that unlike other pure AI products,
02:28it's not just siloed into China
02:30and they're thinking about taking over the world
02:31in a similar way.
02:33It's interesting.
02:34It's much more of a long-term bet
02:35than pure AI stuff is right now.
02:37But I think it shows the company
02:38is thinking quite strategically
02:39about how do they expand worldwide
02:41and how do they find interesting areas
02:43that require a lot of capital
02:45and you build an eventual moat to keep others out.
02:48Do you think that is going to be a big market,
02:50the autonomous driving industry?
02:51Where do you see that going this year?
02:54Yeah, I think it's still early days
02:55that we've said it's ready for autonomous driving
02:58every year for the past five or six years
03:00or it's a year or two away at best.
03:02And now, frankly, we have maybe two cities in California
03:04that have somewhat viable access to this.
03:07And even in LA, you can't take an autonomous car,
03:10self-driving taxi on the highway, for example.
03:12I think it's still a ways off,
03:14but these are long-term bets you have to make as a company
03:16that if and when it does reach a point
03:18of hitting critical mass and acceptance,
03:20there's going to be a lot of cities
03:21where this is going to be a desirable product.
03:23So in my mind, it's a bet these companies need to make
03:26and make strategic sense.
03:28So going back to the bigger picture,
03:30good figures for Baidu this time around.
03:32What might be the hurdles they need to look out for going forward?
03:37I think for all these companies with AI,
03:39the costs are quite staggering up front to build the models,
03:42train the models, support the usage.
03:44So when you have unlimited free tiers,
03:46that of course lends itself to very wide use.
03:48The real question is, can they up the stakes
03:50and increase fees as they go to match the actual usage?
03:54You're seeing this now in the US
03:56with some of the providers,
03:57building OpenAI, for example,
03:59building a programmer level tier
04:01that's $250 a month, for example.
04:04I think all these companies are going to make that evolution
04:06where it's free or cheap in the early days,
04:08then in time they start to monetize
04:09and find the right price point
04:11that both companies or individuals are willing to accept.
04:14And that's still not really shaken out.