00:00This is our interview of the day on technology, dare I say it could be the interview of the third quarter on technology.
00:08Long ago and far away, Charles Phillips was at Colorado Springs and he took computer science before joining the Marine Corps.
00:19What was it like doing computer science with that gorgeous chapel at Colorado Springs at Air Force?
00:26Well, it came along at a time when they were building out the predecessor to the Internet called the ARPANET.
00:33And so being exposed to that, kind of the concept of networked applications and the power of connected computers early on in my career helped me the entire time later.
00:42Did you survive Fortran? I survived Fortran.
00:45Fortran, Algol, Pascal, a lot of different languages.
00:48You really did and I didn't.
00:50Honestly, this is just to get your perspective here on this moment of MAG-7, the accolade of Oracle, full disclosure, folks, Oracle is a sponsor of Bloomberg Surveillance.
01:04You are definitive at Oracle.
01:05Your thoughts on what Mr. Ellison has been doing over the last 18 months?
01:09I think it's fantastic. His vision has always been there and to have the ability to go from one tech trend to the next and take the risk and keep reinventing Oracle.
01:22That's part of his magic. And so that's what's happening now.
01:25Charles, just for our audience, I want to frame your career.
01:29You've been in technology for decades.
01:30I first came to know of you when you were at Morgan Stanley doing some outstanding research.
01:35I understand you've done some things since then.
01:36Talk to us about AI. Can you put AI in perspective of your 30, 40 years of really front lines of global technology?
01:46I would say most of my career, I was automating things around what's called structured data, numbers that could fit in tables, and automating transactions, general ledger, accounts receivable.
01:58What's possible now is 10 times that information that's buried in documents that we couldn't get to before.
02:03Critical information and unstructured data.
02:06And now these things haven't been automated.
02:08So that's what's being possible by AI.
02:11The investment over $360 billion from the hyperscalers this year, 500 if you include enterprises, unprecedented.
02:17The knock-on effects is, you know, depends on who you talk to in Treasury.
02:21That 3% growth, 1% of that is coming from AI build out.
02:24Wow.
02:24And so I think we're at the very early stages now.
02:28We have to get to the second phase of AI.
02:30We're building out the training infrastructure, and that's called inference, using these models to do something practical in business.
02:37Those are the type of companies I'm investing in now that take the technology that Oracle and others are building, OpenAI, Google, and use it in an enterprise to solve a business problem.
02:45And that's called the inference.
02:46So that's going to be 10 times the size of the training market.
02:49Wow.
02:49So what are examples, or what are some of the technologies that you're investing in these days on the inference side?
02:55So we have a company called Cupica based in Uruguay.
02:58They build the agents.
03:00So an agent is something that does something automated on your behalf.
03:03So let's say you want to go travel, and you have to get it approved by your boss.
03:07Your agent can ask your boss's agent, can I travel from Chicago to Paris?
03:11Your boss's agent will look at the cost of that, approve it or not approve it.
03:14Neither one of you have to do anything.
03:15Your boss knows what he will approve.
03:17If you know where you have to go, you can read your calendar and ask that.
03:20So that's what's going to happen, a lot of automation.
03:22Bloomberg Surveillance, a good conversation here this morning with Charles Phillips.
03:26He's co-founder, managing partner, and recognized, but as Paul mentioned, an esteemed career in technology.
03:32I want to know what your advice is to the young Turks doing this.
03:37The crew out of Berkeley at Perplexity is just one example.
03:41Should they sell out to Big Maga?
03:42I mean, I get it, there's a price, but what's your counsel to the PhDs getting it done right now?
03:51Yeah, that's a tough question because each one is slightly different, but I do think that's likely the direction, some consolidation over time.
03:58The competition is good for all consumers and employees to see these big companies going after each other.
04:02It's just accelerating the innovation.
04:04So I want some consolidation, but we're not too soon.
04:06Let's have some more innovation first.
04:08Is Google under threat?
04:10I think Google is underestimated.
04:12They have access to more data than anybody between YouTube, YouTube TV, Gmail, all your search history.
04:18They couldn't get out of their own way two years ago, internally not organized.
04:21Now they are.
04:22If you look at all the leaderboards right now, the nine metrics that you watch on LLMs, they'll lead on six of them.
04:27The number one app in the Apple Store right now.
04:30They are booming right now, so they are approaching open AI at a rate that people don't recognize.
04:35Interesting.
04:36What are some of the – how do we think about the U.S.'s role in AI here?
04:41Is there a role for the government, or do we just want to let the folks out in Silicon Valley and Austin, Texas, and other parts of the world, Uruguay, kind of do their thing?
04:49I think we kind of hands off for a while and let people innovate.
04:52We don't want to be the EU.
04:54I hate to insult anyone, but they over-regulate.
04:56We say good morning to the continent of Europe.
04:58And they miss the innovation because they're so concerned about constraining and regulating as opposed to letting it ride.
05:05And we're more free-form than I think our model works.
05:09Steve Case didn't stop working after a couple transactions, and he has led on distributing technological excellence from sea to shining sea.
05:20How are we doing at that?
05:21Is it just going to be about Austin, Silicon Valley, and three zip codes in Manhattan?
05:27I like what Steve is.
05:28I'm going to invest in his Rise of the Rest Fund for that reason, and he is trying to take new places.
05:32There's smart people everywhere, so we do need to democratize access.
05:36Right.
05:36That's one of the good things about AI.
05:38It lowers the barrier to entry to a lot more jobs, so hopefully everybody benefits over time.
05:49What are you going to do about our electric bills with that?
05:53Just the thermodynamics, just the energy needed to make the Charles Phillips world work.
05:58Well, in a weird way, this is the best thing that could happen to the grid because it's forcing us to invest in it.
06:03We hadn't had any increase in capacity since 2008, and now we're investing in new forms of energy.
06:08So this was the catalyst, and it'll be a little tight for a while, but money's going behind it, and smart people are focused on it.
06:15What is the investing market for technology like these days?
06:19If I have a great new idea, do I go to Sand Hill Road and try to get it funded?
06:24How is the investment environment for supporting some of these new technologies today?
06:28One thing that's changed is you can get a prototype done of a new ID on the application pretty quickly and not be technical because now these coding agents can do a lot for you.
06:37It won't be the final product, but you can get your idea pretty much in some shape to pitch it a little bit more.
06:42So I think you see a lot more sophisticated companies coming with that have an idea that's always formulated a little bit further.
06:48I went back yesterday, and I looked at a paragraph, first paragraph, lead paragraph, from a story that Apple was dead, I think it was April 3rd of this year.
06:57Once again, stock resurrection.
07:00How is Apple doing with all of the acclaim that they've missed the Charles Phillips world?
07:06They did miss a little bit of AI, and that's an issue for them, but they have a lot of users, and everybody's going to be trying to help them.
07:13So whether you work with Google, Anthropics, somebody else, they'll figure it out with some sort of partnership, and it may be the best thing they could do.
07:19It's not their skill set, leverage some of the partnership and get paid for it.
07:23I got 20 seconds.
07:24I need some Charles Phillips wisdom today.
07:26Does Google do a deal with Apple?
07:28Does Apple buy perplexity?
07:30What's the transaction you're focused on?
07:32Hard to predict that sort of transaction, but I think Apple needs to do a deal most likely with Google.
Be the first to comment