[0:00:02] Hi, this is Andrew, and this is Keynote, the daily Now.TV chat show with some of the world's leading thinkers. Hello, everybody. It is Tuesday, August the 29th, 2023. This long, hot summer is coming to an end, but perhaps the long, hot summer of modernity is just beginning to heat up. We've done a number of shows, and I apologize for that, rather.
[0:00:35] Crude metaphor. Done a number of shows on the threat or perhaps the opportunity of synthetic biology quite recently with Ted Anton, Programmable Planet, the synthetic biology revolution. I have to admit, I wasn't particularly convinced by his optimism. Amy Webb also has a new book out on synthetic biology. Meanwhile, of course, the other thing that...
[0:01:02] Is preoccupying tech writers is the threat or perhaps the opportunity of AI? We've done many, many shows, including with Margaret Mitchell, who got fired. One of her, I think to her credit, she got fired at Google for...
[0:01:17] Warning about unethical AI with Gary Marcus, my old friend. He's been on the show many times. He's become one of the leading critics of generative AI and OpenAI. Even with the great English writer, Jeanette Winterson, who, as it happens, is actually more optimistic about the opportunities of AI than one would expect. We haven't done anything, though, on how synthetic biology and AI are coming together
[0:01:47] Coming wave. My guest today, Mustafa Suleiman, doesn't require much of an introduction. He's an iconic figure in Silicon Valley and indeed in London. He was the co-founder of DeepMind, a very successful entrepreneur, worked with Google, now associated with a new AI company called Inflection. And the coming wave is out.
[0:02:09] Next week, I think in the US, middle of the month in the UK. And it's already been included on the long list of the FT books of the year. So already making a lot of waves, Mustafa, before the book comes out. Thanks very much for having me. I'm super excited to be here and talk about the ideas. It feels like now really is the right time. Everything seems to be coming together at this moment. So it's a good time to be alive.
[0:02:38] Most of it coming together or perhaps falling apart. I mean, your book is not particularly optimistic. It's a warning about this dual crisis or perhaps opportunity of synthetic biology and AI. How are they connected in your mind?
[0:02:57] Yeah, I mean, the overriding trend that is not just common to synthetic biology and artificial intelligence, but is actually one that I think we see throughout the history of technologies, is that as technology gets more useful, it gets cheaper because everybody demands it. And we find more efficient and more effective ways of making things.
[0:03:24] That is a trend that has stood the test of time from the invention of the hand axe to the discovery of fire to the arrival of steam and electricity and now to these new technologies from microchips through to synthetic biology and artificial intelligence. And what I've tried to do in the book is consider that idea, first examine whether it is in fact true, and if it is...
[0:03:50] Think about what the consequences are for the future of the nation state and ultimately the future of our species if this new form of power radically proliferates over the next three or four decades. It's an ambitious book.
[0:04:06] Your bibliography includes books you don't normally show up in tech books written by Silicon Valley entrepreneurs like Benedict Anderson's Imagined Community. What's your background? I know you dropped out of Oxford University to become an entrepreneur. How did you go about writing this book? It's your first book. Were you attempting to bring together all your different interests?
[0:04:37] Yeah, I mean, I've always been someone who thinks across a wide range of disciplines. That's always been second nature to me. I'm sort of very lucky in that sense that I'm quite good at teaching myself and I have no problem reading. It's actually, you know, something that I have kept up since university and is a habit that enables me to relax. And so, you know, really the writing process was actually remarkably straightforward.
[0:05:05] Clark, Anthropic, Claude, OpenAI, AI policy, Import AI.
[0:05:32] Actually, the tendency has been towards proliferation and the very few examples where there has been, you know, sort of limited containment and non-proliferation. And so, you know, I was really led by the stories and the anecdotes and the history, you know, which then kind of came together to form this overarching argument about the nature of technology. Some people think, Mustafa, that the Cold War ended, although...
[0:05:58] Some people believe that we have a new one now with China. It's interesting that you choose these words, proliferation and containment. I know containment's an important word in the book. Of course, it was a word that was popularized by George Kennan, describing the Soviet threat and the Soviet Union, the foundations of the first Cold War after the Second World War. Are there analogies between the coming wave, the coming threat of synthetic biology and AI?
[0:06:27] Well, I think to the extent that technology has always enabled states to serve their citizens better or to, you know, go about various kinds of military conquest or advance their nationhood in various ways. I mean, you know, I use the phrase in the book, homo technologicus. What that basically means is that we are nothing as a species, if not.
[0:06:53] Predestined to use tools. That is actually what differentiates us. And every tool, whether it's the glasses that I'm wearing or my ability to pick up a club, enables, accelerates, enhances, amplifies, however you want to describe it, the human condition in various ways. And of course, our tribal form, our nation states, our race, our culture, and so on.
[0:07:20] is the ultimate beneficiary or the driver, both the driver and the product of that tool use in various ways. And so it's not surprising that geopolitical competition isn't just spurred on by these technologies. It's a fundamental motivator of these technologies. And I think there's a range of incentives.
[0:07:44] Not just the sort of political and military, but also the commercial incentives, the desire to tackle hunger and healthcare and deal with a warming climate. There's all these fundamental drivers that lead to everybody wanting access as quickly as possible, everybody wanting to experiment in every which way.
[0:08:09] And on the face of it, that has been an incredible trajectory of progress over the last many, many centuries. And so it's very hard to think about what a containment strategy looks like in that broader context of huge geopolitical incentives. We start to see that now with the new Cold War that people are referring to with China, where quite understandably, they want to get access to cutting edge.
[0:08:38] Satellites, weapons, quantum computers, synthetic biology, AI chips. And of course, just in the last year with the export controls, the US has denied them access to the most advanced chips, which is basically going to stop them from training frontier large language models, the kind of cutting edge of AI. It's interesting, we did a show, Beijing Rules, which is another book that...
[0:09:06] is on the long list. A book about the threat of what the author describes as an authoritarian capitalism. So all this stuff fits together. We're all just sort of playing with all the pieces, Mustafa. In terms of containment and synthetic biology and AI, what exactly are we containing?
[0:09:32] What are we supposed to be thinking about containing? Are we containing the technologies themselves or the companies like Google or perhaps Inflection, your current startup, that own these technologies? I'm sure you're familiar with the work of Ian Bremmer, an old friend of mine. He's been on the show several times. He had a recent piece suggesting that in this new world, the big tech companies need a seat at the table in this post-
[0:10:01] Are we containing the companies themselves or the technologies, or can we not separate the two? Yeah, so Ian's a good friend of mine, and we actually co-authored that piece in Foreign Affairs, calling for the techno-prudential moment. I apologize, yeah, of course. Yeah, and I think it's interesting because...
[0:10:28] This really is a different moment to the arrival of past technologies. These technologies are principally invented and driven by the private actors, the private companies, but they're also widely available in open source. And so the absolute cutting edge is available today, at least in AI.
[0:10:49] to anybody in open source, but it's sort of 18 months to two years behind. That's a remarkable trajectory that something that could cost hundreds of millions of dollars one year could be available a couple of years later entirely for free. And if that trajectory continues, then essentially what we have is more concentrated forms of power. So that is, a single individual has the ability to reach
[0:11:18] A much, much larger group. So it's asymmetric in that sense. But those concentrated units of power are now widely spread. Just as my mobile phone is a concentrated unit of power because it has a million times more processing than it did 40 years ago, which enables me to stream content, record content, create content, and so on. That power to broadcast, which was delivered by the social media age,
[0:11:48] You know, has completely radically changed culture. And we'll see the same thing with AI. The new wave of AI will enable a power to take action, right? Which is a very distinct feature to previous technologies, the ability to do things in the world. You know, your AIs will speak to other AIs. They'll speak to humans. They'll speak to many large groups. They will learn to use APIs, you know, online application programming interfaces, and they'll be able to...
[0:12:16] Make calls right into third party data structures. And that is just a completely different quality of technology to what we've seen before. You talked about the social media age, Mustafa. Many historians, tech historians, see the age of AI coming after the social media age. But are they connected?
[0:12:41] I did a show with Michael Wooldridge, who's the professor of computer science at Oxford. You may know him. I'm sure you know his work, who suggests that the guardrails, the training wheels, so to speak, of AI have been built on our social media content. What's the relationship between social media and the coming wave? Has social media softened us up? Has it taken our eye off the ball?
[0:13:10] No, I mean, I think if anything, it's the reverse, right? I think the chaos of social media has made us all more alert and more concerned about the potential consequences for the AI wave, right? When we went into the social media age, there was a hyper optimism.
[0:13:32] Technology companies were the darlings of the world. They could do no wrong. They didn't need regulating. We bought into the idea that platforms are neutral. The idea that platforms are neutral now, looking back, just seems absurd. Ranking is moderation.
[0:13:51] And the idea that ranking the order in which you see a set of cards on Instagram or a set of search results on Google or Facebook, whatever, the idea that that wasn't influenced by either commercial needs or other kinds of judgment, policy and safety, trust and safety, that now seems to most people pretty obvious. And as a result, this concern that actually outrage,
[0:14:20] Thank you.
[0:14:50] Maybe even the precautionary principle, approaching these things, knowing that they have huge potential to transform our world in very fundamental ways. Second part of the interview, I want to get to your fixes. I know you've got some very specific plans, a 10-step plan, in fact, for dealing with this coming wave. But coming back to the AI issue and the notion of...
[0:15:17] AI being trained on social media. Is there any truth to that, Mustafa? I mean, you're one of the co-founders of DeepMinds. You now have an AI startup, Inflection. Has this new wave of AI, particularly the generative AI driven by the OpenAI revolution, has it learned to think and talk through our social media content?
[0:15:46] Well, not just from social media, but from everything. I mean, there are trillions of words that are included in the training corpuses of models like MyNet Inflection and, you know, many other companies now and open source movements. And, you know, social media content is a small part of it, to be honest with you. It's a very small part of it. Like anything that is available on the open web, you know, has been...
[0:16:12] All kinds of blogs and websites and books, essays, many, many sources. So I think that the other intuition that it's important that people grasp is that the interesting thing over the last couple of years is that the bigger these models get, the more controllable they are. And that's because they're essentially doing an all-to-all.
[0:16:40] connection between all the words that they've been trained on across all the different sources. That's a very crude way of putting it from a technical perspective. But just to give you an intuition, the more computation they have, the more nuance and subtlety and precision they can produce. And that's actually a very good thing when it comes to creating safe AIs and valuable AIs that actually serve the purposes that we're looking for.
[0:17:09] Two years ago, everybody's concern was, well, they're going to be toxic and biased and they're going to produce racist, homophobic, sexist screeds. We'll be flooded with misinformation and we won't know the ways in which we're being tricked and the biases that are included in the model. That's turned out not to be the case, which I think was actually a fairly predictable outcome. It's turned out to be the case that the bigger the models are,
[0:17:38] The more precisely we can align them. My own AI called Pi, which you can find at pi.ai, it's a personal intelligence. It's incredibly safe. It's very, very hard to get it to produce any of these sort of toxic generations. And if you actually...
[0:17:59] Try any of the jailbreaks, the prompt hacks that have been so successful at tricking a lot of the other models like the OpenAI model and so on. They don't work on Py. So we've made technical safety our number one priority and it's sort of demonstrated that if you really focus on it, you can produce a very distinct experience. I would expect you, Mustafa, to say nothing else that your own particular AI is saying. You should try it out.
[0:18:24] Just try it out. Let me know if you... Yeah, we'll have you back on the show and specifically to talk about inflection. But the subject of today's conversation is your new book, The Coming Wave, Technology, Power, and the 21st Century's Greatest Dilemma, this convergence of AI and synthetic biology and the threats and opportunities of tech.
[0:18:47] We're going to take a short break. I want to thank our sponsor, Liberties Quarterly, an excellent publication. We're going to run a short ad, and then we'll be back with Mustafa Suleiman to talk specifically about how exactly we're going to contain all these new technologies and allow them to work for us rather than us working for them. So we'll be back in two seconds. Beyond the news, the noise, there is Nuance Insight.
[0:19:17] Liberties is not just a journal of ideas. It's a meteor of intelligent substance. It's the place to be for engaged citizens. Politics, opinion, substance. Liberties is a triumph for freedom of thought. A quarterly of urgency, of cultural exploration, of intellectual delight, of immaculate prose. It's invaluable. Subscribe now or find Liberties at your favorite bookseller.
[0:19:44] And you can check out more libertiesjournal.com. We are back with Mustafa Solomon, a legendary technologist and now author of The Coming Wave, a book that's already been shortlisted for the financial, longlisted for the Financial Times Business Book of the Year, out next week in the US and middle of September in the UK. So let's get concrete here, Mustafa.
[0:20:10] You have a 10-step plan in the book for quote-unquote containing these new technologies. And one of the things that I think is striking about your argument is you're not falling back on regulation. Regulation may play a role, but you think regulation is old, it's tired, it's perhaps a little bit 20th century. Why can't we rely on regulation? Well, I think the argument is that regulation is not enough. So it's a first step.
[0:20:39] And unfortunately, it won't come first. It will be the, you know, it will be a cultural pressure, social concerns. It will be movements. And ultimately, it will be the industry coming together to, you know, work on self-regulation. You believe in that, self-regulation? You're an insider. Are these companies really capable of self-regulating? They've always talked about it. It never really seems to come out in practice.
[0:21:08] Well, there are precedents here. So the airline safety aviation, for example, emerged as a cross-industry collaboration back in the 50s because they recognized that commercial flight was never going to become a widespread consumer experience unless it was really safe. And so there was a collective sort of self-interest, if you like.
[0:21:33] And, you know, obviously that then became, you know, part of the official regulators and the FAA. But it is, I think, an exemplar case because, you know, you take a look at airline safety now, it's one of the safest modes of transport in the world. And yet on the face of it, it looks, you know, pretty scary being 40,000 feet in a tin tube. And I think a lot of that has to do with the sort of centralized oversight of the communication of failures.
[0:22:00] And safety risks. So the black box recorder, even if there's a minor incident, provides information to the FAA, which then makes a decision about whether to update standards or training practices and shares that information among all competitors. So there is coordination across all of the competing industries. And I buy that. And yesterday, another of the books on Mustafa, on the...
[0:22:28] The long list, the FT is five times faster by Simon Sharp about the politics mostly of climate change. It seems as if the issue of climate change is the other extreme where governments haven't been able to get their hands around it and we can't trust private corporations. Are those the two bookends, the airline industry and climate?
[0:22:50] Yeah, I think that's a pretty good framing. I mean, on the flip side, on climate, it took us 10 years than we expected to reach an international agreement at Paris. I mean, everyone, I think, hoped that that was going to come at Copenhagen in 2009. But that is a seismically complicated challenge, which requires us to forego benefits. And that's the fundamental problem here, which is actually similar to the challenge of AI and synthetic biology. I mean, these are...
[0:23:19] Fundamentally, omni-use technologies. They're general, which means that they are valuable for many, many different types of application. And so the commercial incentive, the social incentive to address tough social problems, not to mention the military incentive, these are going to drive forward development at a breathtaking pace. And so what I frame in the book as the greatest dilemma that we face this century is
[0:23:48] Principally pivoted around that question. It's like, how do you hold back something which is on the face of it so valuable? I mean, this really is going to be the most productive decade in the history of our species if we can get this right. I truly believe that over the next three to four decades, we're headed towards radical abundance. There is no reason why we can't grab all of the benefits here whilst mitigating the downsides.
[0:24:16] Now you're sounding, Mustafa, like a utopian. Do we have to hold our nose, though, on this incentivization? You argued in an interesting way that the Turing test, which, of course, was the test to see whether computers could sound like humans or be as convincing as humans, we couldn't tell the difference, that the new Turing test should see if AI can make a million dollars. Some people might think that rather crap and commercial.
[0:24:45] Especially the anti-capitalists among us. How far can this incentivization go? Do you think bringing out greed in an Ayn Randian sense can result in social good? Well, the purpose of the modern Turing test is to measure what an AI can do, not what it can say. What it can say doesn't necessarily correlate with intelligence, but capabilities, practical...
[0:25:13] Creation of real things in the real world or the use of digital tools. That's something that we really can measure. And over the history of humanity, we have invented a very straightforward and very objective way of measuring value. It doesn't capture everything, but it's clearly the most practical way. And that is via dollars. If somebody is prepared to pay for it in an open market.
[0:25:39] It must have some value. And so the purpose of the modern Turing test wasn't to make a million dollars. It was to provide a very clear, measurable indication of whether an AI system had learned the ability to invent a new product from scratch, to get that manufactured by commissioning it.
[0:26:02] In a factory, negotiate over the blueprints of the product and then actually get it drop shipped and market and sell the new product and actually generate a profit from that. So that's a very complicated series of actions, which is quite simply summed up in the measure of dollars. So, you know, look, I think that we have to be pragmatic.
[0:26:32] And dollars have driven immense progress in our world. And it's an amazing engine of productivity. And I think that once again, this isn't just about being optimistic or pessimistic, utopian or dystopian. It's about just taking a cold, hard look at the facts, identifying the risks and starting to focus on trying to plug those potential holes and mitigate those downsides. And I think that's where the conversation has got to move to next.
[0:26:59] Most of it, one of the striking things about the last 30 years, the world of big tech, these dramatic changes is on the one hand, you've got the appearance of this new aristocracy, guys like you and new centers of power and wealth like Palo Alto where you live. And on the other hand, everyone else seems to have stood still, more and more inequality. In the coming wave and in your analysis, do you address that? Are there ways in which you think?
[0:27:28] A post-nation state world might be able to address this growing inequality, which is, of course, having such problematic, corrosive political and cultural consequences. Yeah, I mean, I think that on the inequality front, I'm a big believer in taxation and redistribution. And I think that we're going to have to get more and more comfortable with corporate taxes going up significantly over the next few decades.
[0:27:59] in the US is taxed at about, on average, 25%, whereas software is only taxed at 5%. And naturally, you tax the things that you want to add friction to, to go slower or to get more costly. And you reduce taxes on the things that you want to be protected. And so taxation isn't just about redistribution of the dollars. It's a political decision.
[0:28:28] That we can use to throttle the rate of change. The time is going to come, and I think we're fast approaching it, where we're going to want to tax capital much more than labor. Because we want there to be time for people to be able to retrain and adapt and improve their skills and continue to compete in the market. Because right now we're turning intelligence into capital.
[0:28:54] And so it's going to have all the characteristics of capital, which is that it can be hoarded and it will compound faster than anything else. So if we want to avoid the amplification of that inequality, the acceleration of the process that's already been underway for 40 years, where there's been stagnant wages for the bottom 60% of people in the US for four decades, then we're going to have to address taxation.
[0:29:21] And I think that we need to lean into that with much more radical proposals than we've got here. And that's going to pale into insignificance to the kind of thing that we're going to have to do in the next four decades. In 30 or 40 years time, it's quite likely that we're going to have machines that do pretty much all tasks more efficiently than humans, apart from maybe the top 10% of complex tasks. So that's going to produce...
[0:29:45] Radical abundance. It really is going to produce way, way, way more for much, much, much cheaper. The question is, how do we create the political decision-making process to tax and redistribute that? And for me personally, my contribution to that is to generate the value and create a business that is open-minded and pro-social and leans into these kinds of considerations. That's why we created Inflection as a public benefit corporation. What's a public benefit corporation?
[0:30:14] A public benefit corporation is one where we have a legal obligation to make decisions that give equal weight to returns to shareholders as to the consequences of people that are not our customers, so the wider stakeholders. So we're trying to internalize the externalities to some extent in our decision-making process, and that's a legal obligation of our commercial charter, of our company charter.
[0:30:44] And it's the obligation of our board of directors. And one of your co-founders is, of course, Reid Hoffman. He's been on the show a few times. I've known him for many years. Do you represent, do you think guys like you and Hoffman, a different kind of Silicon Valley from Hoffman's old friend at Stanford, Teal and Musk and some of these other characters? We did a show yesterday with Jonathan Taplin, who has a new book out suggesting that Teal and...
[0:31:13] Musk and Zuckerberg and Andreessen are ruining the world, these multi-billionaires. So are you an alternative? I'm not sure if you're a billionaire, but you're a wealthy man. Yeah, look, I mean, my goal has never been to make money. You know, the making money is the least interesting. They all say that. Everybody. And I believe it. I mean, when you're that rich, it doesn't matter.
[0:31:38] Well, the second part of my goal has also been to defend the public interest and build the nation state. I mean, if anything, my book is a love letter to the nation state. It's saying, now is the time.
[0:31:55] for us to strengthen it and enhance it in every possible way. Whereas I think some of those people that you've listed, their goal is explicitly to tear down the nation state. They think the nation state is the problem. They think that we would be far better off if we only had the nation state take responsibility for...
[0:32:11] The lightest form of security and that everyone else could establish their own independent libertarian enclaves. I think this is completely insane. And I also think that it's highly unlikely, given what we're about to go through. We thankfully have the democratic process and people are going to feel the increasing pinch.
[0:32:29] of this new wave, and that pressure is going to drive the political process in very fundamentally new ways. We've started to see the tip of the iceberg of that in LA with the writers' strike, which is still ongoing. Yeah, John Taplin talked about that yesterday, and his argument, when we talked about it, was you still need to pay those writers half a million dollars rather than the $30,000, $40,000 that AI is reducing the price to, and I'm not actually convinced.
[0:32:57] Let's end, Mustafa. There's so much to talk about here. You'll have to come back on the show. But let's end with some more concrete ways in which, to quote you, we can strengthen the nation state with this coming wave of new technologies, particularly AI and synthetic biology. You're slightly ambivalent on regulation. You talk about much more aggressive taxes on corporations. What other ways?
[0:33:22] Are you suggesting in your plans for the coming wave that we can strengthen the nation state rather than undermine it in this new age of radically disruptive technologies? I mean, two high level things. The first is that we have to put nation states in charge of auditing and scrutinizing the pace of development of these technologies. And that means that those who are building the technologies, myself included,
[0:33:53] have to proactively make available code access to democratic governments to provide verification of the state of our current progress and the trajectory that we're on. And the other piece of that is that... That sounds rather regulatory to me. Yeah.
[0:34:13] I mean, it will be regulatory, but it will also be pre-regulatory. So we've actually already signed up to do this in the voluntary commitments at the White House. So I visited with President Biden. Can old Joe Biden, can he understand your codes? Well, it's not going to be Joe doing the audits, thankfully. But does he have guys around him? I mean, they're all ex-Google people around him. Can we trust them? Can we trust all of you together?
[0:34:41] Yeah, I mean, I think more than ever, I don't know that it is old Google people, but yeah. Facebook, a lot of, you know, Amazon, Apple, old tech people. Here's the thing, right? I totally agree with the revolving door issue. It is a fundamental risk to the future of the nation state, and it has long been a problem. But here's the flip side. We live in a free labor market, which means that if you're going to go off and pay people a million dollars in the private sector.
[0:35:09] Then you have to figure out a mechanism by which the overwhelming incentive isn't for somebody in public service to do 10 years and then go and spend the rest of their career earning lots of money in the private sector. So we are allergic in the UK particularly, but also in Europe and the US to paying public servants what they deserve. And that means that we're going to have to increase that. I can agree with you more on it.
[0:35:39] Especially in America where public servants don't get any love or respect. So you talked about two big issues. On the one hand, a kind of pre-regulatory regime where big tech has to show its codes to government before it proceeds opening the kimono. What else? What's the other one? Well, so the other thing that we have to do is to make sure that, you know, the cabinet itself has a minister for technology and for AI, right?
[0:36:09] A first-class citizen in the decision-making process, whether it's a CTO or some other kinds of senior technical expert that is alongside world leaders making significant decisions about having technology in-house. So this is the other component, which is that governments have become used to commissioning technology services.
[0:36:37] And you can't really control something that you fundamentally commission. And if you do that for decades, you lose the institutional muscle for actually understanding the detail of what you're commissioning. And so to really have the power to control something and to contain something, you have to have the technical expertise inside, like on the home team, as well as commissioning services. I'm not saying don't use cloud. We've had...
[0:37:03] I think we've even had government CTOs. They've never made any difference. It's not the quality of the individual. It's their power within the cabinet, within the administration. Yeah. Well, I mean, I don't think we've ever had CTOs in the UK government or in much of Europe. And I think what I'm...
[0:37:24] Arguing is that now is the time to invest significantly in having technical expertise on the government side. I mean, that seems to me just an absolute minimum. Well, much to discuss. I'm going to have to get you back on, Mustafa, because there's so many more questions. But congratulations on the new book. It's out next week in the US. I think you're going to New York to launch. And then you're going to be in the UK in the middle of...
[0:37:51] Middle of September for the UK launch. It's on the long list. I expect to get on the short list of VFT. So we'll have you back on the show. Best of luck. I've avoided making any jokes about King Canute in this coming wave, but you're certainly very brave in addressing the coming wave. Mustafa Solomon, honored to have you on the show and we'll definitely have you back in the future. Thank you so much. Thanks a lot. Cheers. Appreciate it.