Tag: Hail

This is an audio transcript of the Tech Tonic podcast episode: ‘Tech in 2025 — All hail the AI revolution’

Reid Hoffman
I see this amazing wave of opportunity. Just about every venture firm is looking at AI ideas. I think we’re going to see any place where human language is used, there will be new products and services given generative AI. That creates a huge amount of kind of start-up opportunity.

[MUSIC PLAYING]

Murad Ahmed
Hello and welcome to Tech Tonic from the Financial Times. I’m Murad Ahmed, the FT’s technology news editor. And in this season of the podcast, we’re looking ahead to what the tech world might look like in 2025. The voice you just heard is Reid Hoffman. He’s the billionaire founder of LinkedIn, a venture capitalist, self-proclaimed AI optimist, and generally speaking, a big name in Silicon Valley. I spoke to him for this episode to find out what the year ahead might hold for the centre of the US tech industry.

Over the last couple of years, Silicon Valley has been taken over by AI fever, with billions being poured into artificial intelligence start-ups and Big Tech groups spending huge amounts on the technology. Of course, the Valley is not the only centre of AI innovation in the world right now. Last week, Chinese company DeepSeek released a cut price AI model that competes with the likes of OpenAI, sending tech stocks tumbling. We’ll be covering the impact of DeepSeek in a future episode. And the arrival of Donald Trump in the White House is shaking up the Valley further, exposing divisions between liberally minded tech workers and the wealthy elite who supported Trump. So if there’s anyone who can tell us what’s going on in the heart of Silicon Valley, it’s Reid Hoffman. So here’s the interview.

[MUSIC PLAYING]

Reid Hoffman, welcome to Tech Tonic. Thank you so much for doing this. Let’s address the elephant in the room straight away. Donald Trump, you were a prominent supporter of his opponent, Kamala Harris. But there were plenty of supporters of Trump in Silicon Valley. What’s the mood now? How do you think this centre of innovation for the world is thinking about what is coming under the Trump presidency?

Reid Hoffman
So I think, you know, generally speaking, Silicon Valley tends to be very optimistic. If there is a religion of the area, it’s that we can build technology to make a massive difference to, you know, humanity, to the human condition, to society, to industries, to products and services. You know, the whole way. And I think that the ability to build amazing new technology companies is still there. There were many more people who were supporting Vice-President Harris than President Trump, but there were obviously some significant supporters of President Trump. And I think that the general view is that the ability to innovate, the ability to create things, the ability to do things will still be there in full, just as we kind of in the modern age live in a technologically accelerating age. And so I’d say it’s bullish on the ability to be kind of technologically forward, with some uncertainty as to what kinds of possible chaos or other uncertainties that might get introduced.

Murad Ahmed
Forgive us on focusing on the chaos straight away, but you wrote in an opinion piece for the FT your hope for the tech industry and the potential growth under Trump. You are, after all, you know, chief optimist for Silicon Valley, I think. But you also wrote that the Trump administration would adopt a mercantilist AI policy in particular. What did you mean by that?

Reid Hoffman
So one of the things is, you know, you listen to President Trump’s campaign, it’s kind of a, you know, America first, second and third versus kind of, generally speaking, a multilateral approach. A promise of tariffs, which, you know, most businesspeople, most Silicon Valley people think are, generally speaking, not a great idea because they introduce lots of various forms of friction, trade war, you know, potential favouritism in how they’re implemented. You know, and in particular, I actually think it’s very important within AI technologies to be kind of inclusive within what is frequently referred to as western democracy, where we have human rights and individual values and limited powers of government. And I think that’s one of the reasons why it’s important, I think, to be multilateralist. But, you know, if you listen to the campaign promises, it’s exactly the reverse within the Trump campaign.

Murad Ahmed
You mentioned AI. AI crosses borders. It’s going to be hard to control within any particular borders, it’s the nature of the technology. But you also point to some of the dynamics of some of the key players that have been crucial to Silicon Valley and American innovation of the past few years. One of your worries is about Elon Musk, the founder of xAI, an AI start-up, as well as the chief executive of Tesla and others in particular. I wonder what you thought about some of the personal dynamics that are occurring in the Valley right now. For example, Musk has a particular animus against Sam Altman, the chief executive of OpenAI, which has been a leading AI start-up at the moment. Are you worried that that animus is going to play out in Trump’s AI policy going forward?

Reid Hoffman
Well, I certainly worry that it might. You know how in a good universe should a government be functioning? It should be functioning to say, hey, we want every healthy American institution to continue to prosper, including an entire industry, a whole set of businesses. We want them to be able to deliver the products and services they can for the industry, for society, for consumers, for employees, for investors, and to have an industry-wide point of view versus a . . . only what’s good for me or only what’s good for my companies or only what’s good for my interest, you know, kind of point of view. And I think it’s very important for American society and for the world to have that kind of, you know, much more broad perspective. And the fact that I think Elon called Sam evil. You know, he’s named me in his lawsuit, which is kind of, you know, everything in the kitchen cabinet.

Murad Ahmed
I don’t want to immediately bring up the lawsuit. But yeah, I mean, he has his targets, clearly.

Reid Hoffman
Yes. And so, you know, obviously integrity and character, I would say, look, since I’m involved in these kinds of lawsuits and so forth, I should keep myself distinct from the operation of government in these things. And yet you see all kinds of weird boundaries. Just like, for example, you have the UK asking Elon to come over and talk about some of where X.com is involved in inflaming racial hatred or other kinds of violent events. And his reply is, we’re going to call ministers over to answer for their denial of American freedom of speech. You know, like, well, you know, there are two different countries and countries have the moral legitimacy of their own governments and so forth. And you don’t call other governments where you’re not even a government official to be saying that. And so that kind of crossing of boundary portends potentially kind of dangerous myopia and dangerous conflicts of interest. And that’s the thing you have to be very careful about.

Murad Ahmed
If these conflicts of interest were to play out in a Trump administration, what is the actual policy that could be implemented that could allocate winners in the future development of AI?

Reid Hoffman
Well, I don’t want to speculate too specifically. I hope to be wrong about this worry. I would be delighted if this worry were 100 per cent incorrect. But, you know, like, for example, if you combine it with tariffs, I mean, it’s kind of simple. You say, well, OK, everything that isn’t xAI has tariff penalties and xAI doesn’t. I mean, that’s a little bit, hopefully, hyperbolic. But, you know, you can get the questions of, all right, so xAI’s coming from behind, so let’s try to slow down all of the leaders, that xAI can catch up. You could just do all of that kind of thing as ways if you’re implementing government policy to try to privilege one company over others. And I think that’s a very, frankly, very destructive thing to do. It’s destructive for the industry. It’s destructive for American society. And, you know, hopefully those don’t happen.

Murad Ahmed
You also brought up the idea that one of the areas that we should have a look out for is the crypto industry. I mean, it sounds like in some ways you’re bullish that less regulation of the crypto industry could be a good thing. But you’re also worried that, again, Trump may end up picking favourites in this industry. We’ve talked about Elon Musk conflicts of interest, but Trump here has his own conflicts of interests. Given your sense of where the Valley sits at the moment, how likely do you think that is that actually, it’s not just that we’ll have a big bang moment for the crypto industry; that it will actually benefit some over others.

Reid Hoffman
Well, I don’t know if the Valley has gotten to that concern yet. I think that a lot of the crypto people in the Valley are happy that they think that the previous kind of uneven, unclear rules, they’re happy that it’s back to an innovation-forward approach because very likely, under the next administration, there will be kind of a, hey, you go out and build whatever kind of crypto things you can. The concern is that, you know, obviously when you have a, you know, kind of a president who has their own very deep personal economic interests, you know, whether it’s a, you know, social media platform Truth Social, whether it’s a crypto coin, you know, the FREEdom coin and other kinds of things. This creates an entire surface of not just, you know, obviously conflicts of interest, but also ways that international oligarchs can do bribery by propping up stock prices or buying crypto coins or other kinds of things. And those can lead to some very unique and very present concerns about direct conflicts of interest with the American people.

Murad Ahmed
And you also, Reid, I mean, you pointed out that the Musk lawsuit against OpenAI, you were named, I mean, you are and have been a prominent supporter of the Democrats. There must be an underlying fear for you personally, but other people who’ve been democratically supporting other people for president, that there could be retribution in some ways. Is that something that’s crossed your mind?

Reid Hoffman
I think that’s crossed the mind of many Democratic supporters. You know, some of Trump’s rallies are going, they’re gonna after the enemies within. There was obviously a lot of invented fiction, from immigrants eating dogs and cats to everything else. And you know, a lot of essentially lies and slander and, you know, promotion of hatred. So given that, I’d say that there are rational concerns and I have them, too. But I would say that if we wanted to be promoting a kind of a healthy country would be kind of how do we build bridges to building the future. But obviously, when you look at the campaign trail, that doesn’t seem to lead in that direction. So, again, I have hope, but I also have deep concerns.

Murad Ahmed
And let’s talk about your hope, because I’m sure there will be people listening to this thinking the markets have very much rallied ever since Trump’s election. There is positivity out there in the business community about his presidency. You also have said that you hope that Trump could be great for American innovation. He could be great for the technology industry. To frame it in a way that you have in the past, you know, what could possibly go right with a Trump presidency?

Reid Hoffman
So, you know, one of the things that I think is actually very important for all world leaders, but especially American presidents, is to have a kind of business-forward point of view because, you know, business is how we create the prosperity of society. It’s how, you know, kind of aligns all the work that people do to create products and services and infrastructure. Government kind of lives on the taxation revenue that comes off this. And so when you get to kind of these considerations of like, for example, what should regulation be, frequently people say, well, the only real question of regulation is to prevent bad accesses from, say, businesses. And by the way, that’s of course, a very good thing to happen. But it’s kind of a more complex function because you want businesses to also innovate. You want businesses to create new products and services, new jobs. And if you’re . . . if the entire function is only let’s make sure absolutely nothing goes wrong, then very much less can go right. And so it’s kind of this balance that you’re striking and trying to regulate only in the areas that on balance, the harms are worse than the possible benefits.

And so, for example, one of the things that Trump did in his first term was to say, hey, if you want a new regulation, remove two, right? Because we need to be refactoring these regulations to be allowing that innovation. We need to be giving business the channels and freedom to explore areas that we don’t know yet. And that’s the kind of thing that I think we’re going to see from a Trump presidency that will be very good for business.

I also think, for example, when you get to kind of AI and the power needs, you go, OK, well, we need to refactor regulation around the creation of power plants and data centres to enable that to function well. And I think the kind of regulation bulldozing that a Trump administration does could be very helpful for that. And so those are the kinds of areas that I think can be very positive for business. And by the way, markets basically almost always rally when a leader comes in saying, you know, I’m going to be business-friendly, I’m going to reduce taxes more, more dollars to invest. I’m going to reduce regulation. You have more freedom to innovate and more ability to take risks in order to build your business. And in anticipation of that, markets almost always go up.

To keep markets up, we’re going to have kind of more stability in the overall system, more of an effort to create unity and harmony in the country and in the industry than chaos and division. And so I think those could be future negative effects. But I think that that’s part of the market rally.

[MUSIC PLAYING]

Murad Ahmed
Moving on more broadly on what the Valley will look like in the year ahead, can you paint a picture of what you think Silicon Valley, the US tech industry, is going to be like in the coming year?

Reid Hoffman
You know, I think Silicon Valley is kind of this . . . think of it as almost like ideas and technology and business factory and power plant. And so I think a lot of the way that Silicon Valley will look next year is how looks this year and the year before, which is a bunch of folks who are looking at what kind of opportunities are being created by new technologies and what new businesses could be created and trying to get those created and getting those deployed. And then as companies get to scale, how do you scale them, you know, across the entire industry, across the entire world? And I think that will be the same. I think, you know, probably well see some pick-up in the crypto industry because of the change in the regulatory stance. You know, obviously, we’re in in full motion on AI. You know, from start-up to massive company, everybody’s working on it. And what is the various kind of revolutions that AI can bring? And so, you know, I think that we will continue to see all that.

Now, that part of it is we don’t know, with all of the uncertainty and idiosyncrasies that could emerge with the next administration, I think that’s the default path. Right? I think the question comes in what in actually what happens in not just global events, but also what the next administration ends up doing.

Murad Ahmed
It is incredibly hard to predict. Equally, you’ve made a career out of making the right predictions and following your nose to where the innovation economy is taking us. And one of those ways is artificial intelligence. You’re a co-founder of Inflection, and it’s something that you have thought about a lot. Just in terms, again, about the Valley, because we have listeners who think and read and hear lots of things about how this area of the world is centre of innovation in the world. Could you describe how AI is transforming entrepreneurship and technology in Silicon Valley and what we might again see coming down the track?

Reid Hoffman
So when you have this major new technological revolution, a massive number of opportunities open up for start-ups. And so there are start-ups all over the place. There’s start-ups like Sierra in customer service, there’s start-ups in sales, you know, Tome, others. There are just start-ups all over the place and a whole bunch of different coding assistants. And I think we’re going to see any place where human language is used, there will be new products and services given generative AI, and so that creates a huge amount of start-up opportunity. And our frequent speculation is, well, how much opportunity is going to be captured in start-ups and how much is going to be captured by the large tech companies? And the answer is going to be so much opportunity, the percentage doesn’t really matter because it’s doing a lot captured on both sides because of the levels of innovation.

Now, for the next year or two, I think we’re going to see a lot of things that are kind of the straight line from kind of what we see with ChatGPT now and where a set of different tasks that you do using language will be amplified. And we’ll have a ton of co-pilots. I think actually we’ll have many, many more co-pilots than we’ll have, as it were, person replacements. And roughly speaking, I think a heuristic is when do you naturally see a person replacement is when you have a job that is essentially a human being trying to act like a robot, like following a script, like customer service and where you have co-pilots is when you know you don’t actually have the job, which is kind of just following a script.

And I think that this year, in 2025, we’ll see AI every place that you’re seeing the use of language, and here I mean language very broadly. So it’s not just obviously in meetings where you have note-taking and summaries and all the rest, but also of course, financial analysis or coding or legal or medical diagnosis or other areas. And I think you’re going to see co-pilots across all of this. And I think the majority of stuff you’re going to see, you’re going to see from start-ups. Because start-ups do the you know, like, OK, well, how does this work here? How is this basket integrated to an individual’s life or an individual’s work here? What’s the specific kinds of products and services? What’s the go-to market strategy? And part of the reason why so much opportunity comes out of start-ups is when you have thousands of start-ups looking at this, taking risk, being bold.

The large companies will, of course, realise great opportunity, not just through their clouds but through some of their products and services as well. But there’s a limited number of things that they can focus on. So that’s the reason why I see this amazing wave of opportunity. And that’s part of the reason why basically, if you look at what’s going on in Silicon Valley, just about every venture firm is looking at AI ideas. The vast majority of start-ups are pitching themselves as AI ideas, even when sometimes the AI is a little bit like buzzword bingo. So a software company. No, it’s an AI company. Right? So anyway, so that’s . . . 

Murad Ahmed
There’s a lot of start-ups with .ai at the end of them at the moment.

Reid Hoffman
Yes, exactly.

Murad Ahmed
To that I think there’s been loads of interest, some might say hype about how AI might transform things. It’s clear that it really is having a transformative effect. And what you’re describing is in the next wave of companies that we might see, there are applications rather than the kind of the platform, the model say OpenAI, the race between model companies OpenAI, Google and others. Now we might see, you know, hundreds, thousands of start-ups trying to build on top of these platforms, trying to build applications.

But it sounds like in this year to come, what we’re looking at is companies that are working around co-pilots, you know, assistants to human beings rather than something a bit more fully fledged that can take over human roles. I was wondering whether you thought that that was a kind of a stable trend or actually AI might develop much more quickly and be able to do things that human beings have typically been able to do through reasoning and other aspects?

Reid Hoffman
Well, I think what we certainly know is that where there’s places where we’ve essentially been trying to do to make humans act like robots, following a customer service script, etc, all of those areas will be massively transformed in terms of how AI comes in, because they will do that better, cheaper, faster. I mean, customer service is a major cost for all companies, so they try to avoid getting cold as much as possible. Like in the US, it tends to be between $10 and $35 per call. So it’s like very expensive. And yet if you go, hey, we have an AI that’s running. It can say, hey, call, talk to me. What’s is . . . are you angry with this, what’s going on with the product? Are you sad? Or can I help you with that and so forth. And that might all, you know, kind of really elevate the entire human experience in addition to shifting the job.

Now, how far do these co-pilots go? Because when you get to major swaths of the rest of the work where we value our human agency based on how we act, think and decide and reason, and we say, well, one thing that these AI agents are definitely showing is they’re showing the ability to reason in some ways, their ability to plan, their ability to, you know, take notes, the ability to write an analytic summary. That exists already today. And as part of that, does that mean that as opposed to having, you know, Susan or Bob be the engineer, we just have the AI agent be the engineer? And we don’t know exactly how far that will go.

I tend to think that the technologists and the accelerationists tend to go, it’s going to get there really quickly because look at how fast these things are improving. And, you know, when I kind of look at my own use of these AI tools, I tend to go, well, OK, yeah, it makes things a lot quicker for me. It makes me a lot more productive. But I still have a variety of ways that my acting in partnership in co-piloting with it brings out a better, higher, important function. I bring some context-awareness, I bring judgment, I bring an awareness of is this actually solving the problem? Like one of the examples that, you know, you look across all AI agents today, including their history of progression, whether it was GPT 3, 3.5, 4, GPT 01, they don’t realise when they’re making an error.

Small example, you pose a prime number problem. They get it wrong. You say you got that wrong. Then it’s, oh, yeah, sorry I got that wrong. Here’s the answers. Well, that’s wrong. Sorry, I got that wrong. Here’s the answer. No, that’s wrong. They don’t do what a human being does, which is clearly I’m not understanding this. Right? And that’s the kind of thing that leads me to have, kind of, call it informed and reasoned hope, and that even with massive increases in the capabilities of these systems, the person-co-pilot combination will still be so much more performative in all of these cases. You know, engineering, in medicine and law, in reasoning, in meetings and in making decisions, doing analyses. And then if you just say, well, wait a minute. So if you accelerate me so that I am 10 times more productive, does that mean that we have 10 times fewer jobs? And actually, in fact, I think the answer with productivity is no. No, it means we have 10 times the number of different things we’re doing, in creating and so forth.

And so like, for example, if you say, hey, I make salespeople 10 times more productive, you don’t go, oh, I’d rather have fewer salespeople now. It’s like, No, no, I’d rather sell 10 times more things. So I think that’s the reason why I’m optimistic about this future. But like anyone who says with certainty, they know exactly what it is, they know that there’s going to be a whole bunch of job loss and replacement. Actually, that certainty doesn’t exist. Anyone who says there certainly won’t be, that certainty doesn’t exist. And I’ve obviously people with uncertainty get nervous. But when we look at the history of technological transformations, the transition can be very rough and we should go into that. But we end up with an enormously productive and more generative and more supportive of kind of a broad-based prosperity than we did before. And I think that’s what’s possible. You know, what could possibly go right? And that’s what we should be working towards.

Murad Ahmed
I think your hope and optimism, not only in this conversation, really comes through in the book you’ve just written called, Superagency. I saw it as very much the utopian Silicon Valley view on AI. What makes you so optimistic that AI will actually benefit all of us, all of humanity?

Reid Hoffman
So, you know, part of what I do is I argue that, look, in all human history, this is actually what’s happened. When the printing press comes out, there’s a discourse that this destroys human society. This destroys the role of experts, in this case, the priests and, you know, causes misinformation, causes the degradation of human cognition and so on throughout the March history. And in each of these cases, when you get to the other side where we’ve integrated the technology in our society, you end up with a far better robust human amplification of technology giving us superpowers, you know, hence super agency. There’s a historical generalisation.

Then you say, well, is this time new? Is it new because it’s much faster? Is it new because it’s cognitive, not just physical? Or is it new because it’s agentic technology? It has, you know, it’s imbued with various forms of agency. You know, I actually believe that even if we didn’t intervene specifically to try to make it into this kind of co-piloting human amplification, it would naturally end up there. But of course, that doesn’t mean we shouldn’t try to intervene and to articulate what our design principles are. And part of my optimism is that, you know what people don’t realise, which is like companies are not just regulated by governments, they’re regulated by customer response, they’re regulated by employees, they’re regulated by investors, they’re regulated by the communities that they’re involved in, the response to the press.

So there’s lots of things which kind of shape them, trying to offer things that are positive for society and positive for the long-term brand and positive for the revenue line. All of those things are kind of natural feedback loops. And that’s part of the reason why I’m intrinsically optimistic and intrinsically positive on this. I describe myself as a bloomer versus a doomer or a gloomer. But that being said, that doesn’t mean, oh everyone should just sit back, take a siesta, drink a glass of wine. You know, no worries here. No, there’s lots of terrain to navigate. There’s lots of potential potholes and accidents and mis-steps and transitions can be difficult. So going in with the intentional practice of how do we help elevate humanity and bring us all along and give us the superagency collectively and integrate in our society well, to take a positive design stance is actually, of course, very important.

Murad Ahmed
One of the many ways you’re so well known in Silicon Valley as an investor, as a venture capitalist. And I wonder what you think in 2025 investing looks like? Not necessarily just in AI, but just investing in tech. And put a more specific question for you, where will all the money come from? How will this be paid for? I mean, if we look at the kind of the billions of dollars that have gone in so far to build AI models, who could possibly pour more billions into, for example, OpenAI? Is that even a possibility? Are you an optimist that investing in companies like that across the board can really grow from here?

Reid Hoffman
Short answer, yes. The billions that have invested, there’s many more billions of capital. The fact that this is such a massive — you know, I call this the cognitive industrial revolution — such a massive set of economic opportunities that I think the capital is there. I think, you know, a little bit of like what we’ve seen last year, maybe we’ll continue to see this year as well. Maybe there’s all these expenses, but where’s the revenue? Because, you know, people like to have the drama story because it’s like, well, you know, ideally you put in a dollar today and then tomorrow you have the revenue shows. Now, this part of investment, as you put it in the dollar today and three to five years, you have a whole business and industry. Or seven years, you have a whole business industry that has substantive revenue. And I think that’s what we’re going to be seeing with AI. And so I’m bullish on the . . . as an investor.

Murad Ahmed
Reid Hoffman, thank you so much for your time.

Reid Hoffman
Thank you. It was a pleasure being here.

[MUSIC PLAYING]

Murad Ahmed
That’s it for this episode of Tech Tonic. Thank you to Reid Hoffman for speaking with me. We’ll have another episode soon.

[MUSIC PLAYING]

Tech Tonic is presented by me, Murad Ahmed. Our senior producer is Edward Lane and our producer is Persis Love. Executive producer is Manuela Saragosa. Sound design by Breen Turner, Sam Giovinco and Joe Salcedo. Music by Metaphor Music. Our global head of audio is Cheryl Brumley.