About Paul Johnston
Paul Johnston is an interim CTO, CTO and strategist who has particular interests in serverless, cloud, startups and climate change. Formerly, Paul served as a Senior Developer Advocate at AWS for Serverless and CTO of multiple startups, including one of the world’s first serverless startups. Paul’s also a keynote speaker, tweets a lot at @PaulDJohnston, and blogs a lot on Medium. Right now, he may also be working in stealth mode on something (it’s probably serverless)…Links ReferencedTranscript
Speaker 1: Hello, and welcome to Screaming in the Cloud with your host, cloud economists, Corey Quinn. This weekly show features conversations with people doing interesting work in the world of cloud, thoughtful commentary on this state of the technical world, and ridiculous titles for which Corey refuses to apologize. This is Screaming in the Cloud.
Corey Quinn: This episode of Screaming in the Cloud has been sponsored by Manifold
, Manifold powers marketplace infrastructure that connects millions of developers to the best APIs, tools, and services in the fastest growing communities, and also Kubernetes. They offer a complete toolkit that allows you to deliver your API first product to millions of developers. Check them out at manifold.co
. Again, that's manifold.co
. Thank you for sponsoring this ridiculous podcast.
Corey Quinn: Welcome to Screaming in the Cloud. I'm Corey Quinn. I'm joined this week by Paul Johnston. Paul, welcome to the show.
Paul Johnston: Hello. Hi. Very nice to be here.
Corey Quinn: So once upon a time you were an interim CTO, CTO and strategist, and you were focused very heavily on Serverless. In fact, when I first met you, you were a senior developer advocates at AWS for Serverless.
Paul Johnston: Yep, that's correct. It was an exciting time at AWS, and there were only two of us at the time, myself and Chris Munns. Yeah, it was an awful lot to do, and Lambda was a little bit younger back then. There were an awful lot of companies to get around, an awful lot of organizations to talk to, and I was based in [inaudible 00:02:07
] Chris was based in North America. So yeah, it was a fascinating time to be around AWS back then.
Corey Quinn: It certainly seems so. Lately though, it seems that you haven't been speaking nearly as much about Serverless as you have about climate change. And what makes you interesting is that I can find any number of random yahoos to come on this show and talk about Serverless in a variety of different tones of voice. But no one in our space seems to be talking in any meaningful way about climate change except for you. And based on conversations that we've had, I know you to be a thoughtful, sincere person and you're not a ridiculous crank. So let's talk this half hour about climate change and how it impacts cloud computing.
Paul Johnston: Well, thank you very much for the eh saying I'm not a crank. I hope I'm not so. It'd be a great thing to talk about, I think.
Corey Quinn: About climate change or being a crank?
Paul Johnston: Haha. About climate change. Let's talk about that. Hopefully, I'll come out a lot better at the end of it, and not a crank.
Corey Quinn: So you've given a presentation on this. You've written on Medium, which is fussily named because it is neither rare nor well done, and you wrote a white paper on energy usage in data centers contrasted between the large providers. How are we doing from a climate change point of view in the cloud computing industry?
Paul Johnston: So the cloud computing industry, it's not doing too brilliantly, although there are some real bright spots and some real low spots. The industry itself is ... If you look at the whole data center industry, and there are different opinions on this and it's difficult to look at, but the whole industry appears to be actually a pretty bad pollution in terms of greenhouse gas emissions. So some will say it's below 1%, others will say it's 3%, 4%, 5%. I take a roughly middle view of about 2% of greenhouse gas emissions are worldwide. This is our down to data centers.
Paul Johnston: Now, if you think about the fact that actually we're all streaming Netflix and we've got all of these things stored in the cloud and we've got all of this storage and we're using cloud providers, there's 2% of ... one 50th of the greenhouse gas emissions that are going on in the world are basically down to us using computers as techies. Then you look at where the growth is in the market, and the growth is in cloud computing sector. It's, yes, we've still got all of these corporate data centers but they are all beginning to move, and over time we'll move and shift to cloud computing.
Paul Johnston: So the the fact is that we've got to look at cloud computing and we've got to look at cloud computing as, is it a good example of ... is it good at looking after resources and looking at whether or not it's emitting well or badly, or what what's it doing? Is it offsetting? Is it doing all of these things? So myself and Curry created a white paper last year. We did a little bit of research and we took the top six cloud providers, and we looked at them and we basically said, "Who's doing well? Who's doing badly? What are they doing, and how are they doing it?"
Paul Johnston: We effectively came up with ... and just take the top three because it's the easiest. The top three of AWS, Microsoft Azure, and Google. Google are the exemplar. They are doing really well. If you have a look at their sustainability page on Google Cloud, they try to offset all of their usage with renewable energy, and they do a really good job of it. It's not perfect. So if you see their China usage, you will find that they are struggling to find renewable resources to offset, but they do as much as they possibly can.
Paul Johnston: It offset in other areas of the world to do that. But they do a very good job and they offset a 100% of their usage as much as possibly can with local renewable resources, but otherwise by renewable energy credits. Microsoft are, I think if I remember rightly, over 50% now. I think over 60% possibly, and then buying renewable energy credits for the rest. But Amazon's AWS is the one that we're all concerned by, and at the moment, they are stating 50% and they're not actually saying what what they are now. They said that in the beginning of 2018, and they've grown since then. We don't know what their electricity usage is now. They don't release numbers.
Paul Johnston: So we're not getting updated information and we're reaching the point where we're constantly asking, "Let's have some updated information, let's have some updated information," and they're not giving it to us. But they have four regions which are Oregon, Montreal, Ireland, and Frankfurt, and those regions are a 100% sustainable. So if you can put workloads that are non production or not needed all the time, then put them in there and you're 100% sustainable, you can be green and still use the cloud and still use AWS. But otherwise, us-east-1 is basically, just don't use it if you can at all avoid it, for environmental reasons if nothing else.
Corey Quinn: There have been many reasons not to use that particular region, but this is the first time I've heard this one.
Paul Johnston: They are very, very many, and it's it’s basically a very simple thing to consider. We've looked at things like machine learning workloads. You could move those out. You don't need to run those in us-east-1. You can move them anywhere. Just all of these kinds of things. We just automatically use us-east-1. Well, I used to work for AWS and there are a lot of people there that I really ... I have a lot of fondness for, and I remember, a lot of my times there.
Paul Johnston: Well, you know it's still a big company. It's got a lot of money. It could buy some renewable energy credits to at least offset to 100%, and get there and then start investing in the renewable infrastructure. It could do these things, and actually make a big difference in the world, but it appears to not be doing that, and we're not entirely certain why.
Corey Quinn: So other than I guess people who can think further ahead than next quarter, there are a number of people who are clamoring about climate change across the spectrum. Why don't we see that in tech? Or if we do, we have probably people in tech who are sounding the alarm on this, but they're not talking about cloud computing. Why?
Paul Johnston: I'm not sure. I think we've got a bit of a problem within technology for a number of reasons. But I think we see a technology as not really a part of the problem. So one of the things that I've found, looking into this, is that technologists don't really see technology as a whole as creating a carbon footprint. One of the CTOs that I've talked to recently said to me, "I don't know why you're talking about all of this stuff. My company doesn't have a very big carbon footprint," and this was a company that has a huge data center.
Paul Johnston: I was looking at him, and just trying to understand how he didn't see the correlation between the amount of electricity he uses in his data center and the power stations that create that electricity, and it's because in our world ... There are a lot of conversations that go along with this and I'm not trying to agree with any political side. But in our world, because we have created this globalized world, we've actually abstracted away the carbon and abstracted away the emissions so far away from the turning on of the light or the turning on of the computer and the use of our electricity to the point where we don't see the impacts.
Paul Johnston: That's especially true for the culture that we live in. So when you talk about the climate impacts, there's a lot of ... I mean this is true for technology, but it's true for the culture that we live in in terms of the West. So when you talk to climate activists, what they all talk about is the Global South and the Global North, because actually the impacts of climate change is being felt in the Global South and they are being felt now, just because of the 1.1 degrees we've already had of warming. Whereas, the Global North doesn't feel those changes quite so much. Even though we've had a few hurricanes and they've been really bad, we don't feel them, especially in the UK and North America.
Paul Johnston: We haven't felt them anywhere near as badly as somewhere like India or Africa has felt them in terms of droughts or in terms of extreme weather events. So we've got this, we’ve got this because it isn't my problem, it hasn't affected me, and because I don't see the problems that are coming down the road, and everything's abstracted away, and all these things don't really matter. I think we have a problem in tech in terms of taking responsibility, and in terms of how do you say I am responsible for something and taking that responsibility forward. I think we've really got this problem with convenience and responsibility there.
Corey Quinn: Do you think that this hits on the entire idea ... I guess, we'll go back to Serverless for a minute, of sure there are servers there, but I don't have to think about them or care about them. If we take a look at cloud computing, there are things that I obviously have to care about. We can pervert the shared responsibility model, as an example here. I don't have to care, for example, about what hard drive vendor AWS uses. That is so far below the level of things I need to care about.
Corey Quinn: It feels deceptively compelling to be able to say, "I don't care where they buy their power, because I shouldn't need to." That’s it feels like going down that path is the only realistic stance someone can take because they're not empowered to directly change it by most reasonable perspectives that I've heard, and they're trying to solve business problems that don't directly align. This has been something that gets a surprising amount of pass, and I don't think it's accurate, but I suspect that might be where that attitude comes from.
Paul Johnston: Yeah, and I think I would agree. I think you know since the cloud computing ... the convenience of it, the convenience of being able to click a button or even just send an API call, and you have this huge amount of computing power at your fingertips is a world away from having to provision or even buy servers, and then provision them and set them up and make sure you've bought your electricity, and then you set up things called power purchase agreements.
Paul Johnston: You know have to buy power from a power company. You know that’s the kind it's so far removed from all of that. You abstract all that complexity and you have this convenient service that gives you all this compute power. It's so far removed. Then I can completely understand how a technologist would feel disempowered in that scenario to say to someone like AWS, "Oh yeah, you need to change your power usage. You need to change the way you do power. You need to change the way you think about your electricity."
Paul Johnston: And one of the CTOs I spoke to about this recently said, "Well, why don't we just get everyone to move to Google?" I said, "Well, how's that going to work?" One company moving to Google is not going to make one you know... an AWS, or you know if we're talking about Oracle, for example. Let's not just demonize AWS here. I'm not trying to do that. I'm just using them very much as an example. Oracle's only at 33%, for example, in terms of renewable energy usage.
Paul Johnston: You can't just say, "Well, one company moving to Google," which is the gold standard here. You can't just say, "Let's move everyone to there," because one company moving will make no difference. We've got to have awareness across the organizations who are using cloud, and then get them to combine and get the industry to understand itself. This is what activism is. It’s not activism in this scenario is not about getting everyone to vote with their feet and move away.
Paul Johnston: It's about actually getting the industry to realize its responsibilities, and start to have the conversation and create the conversation. And I think the industry should be having the conversation that says, "Right. I think we need to do something about this. I think our responsibility is to everybody in the world," and this this industry is huge. It's got a huge amount of money involved.
Paul Johnston: And actually we should care about the fact that our cloud provider is is providing all this convenience but it’s not convenience without a cost. And that cost shouldn’t be at the expense of our childrens’ future And it shouldn’t be at the expense of you know cities in India in 2050 being unlivable. I mean essentially that’s what we’re talking about is partly and while I know that there are people going well that’s not me using cloud computing is not going to do them. I understand that one person doing one thing is not going to change it. This is about changing perceptions across the industry.
Corey Quinn: The other challenge is that people have a sense of immediacy, where I wind up talking to people who are big in the cryptocurrency space and they have the temerity to yell at people who travel too much of, "Well, you understand that the jetliners have a massive environmental impact." Yet, what do you think mining your Dunning Kruger is causing to the environment? It's one of those incredibly wasteful things, it almost feels like it's too surreal to exist, but it does.
Paul Johnston: Yeah. And I think the problem is that ... I mean, you look at the Bitcoin specifically. Bitcoin and Ethereum have a proof of work, and proof of work is essentially a massive lottery that everyone takes part in. And if you take part in the lottery, you use compute power to try and guess a number to win some Bitcoin or Ethereum. That's essentially what it is, and everyone joins in this lottery and they win some money at the end of the day, and then 10 minutes later they get to do it again.
Paul Johnston: Huge amount of electricity to win some Bitcoin, and the the research is showing ... I think I did it recently. The research showed something like driving a car 800 miles is the same as one Bitcoin transaction, and that in terms of global emissions. So when you start talking about it in those terms, you start to really get scared about, is this really worth it? And I think we have a real problem in terms of when we start to point fingers with somebody's individual behavior. An individual's behavior is actually really not that important in the grand scheme of things, but when we gain together, when we do something together as an industry, we will make a difference.
Paul Johnston: Things like basically saying, "Proof of work is a terrible thing to do. We should not be doing that," and saying, "Yes, our cloud computing should be 100% offset, and we should be investing in data centers that are completely renewably powered and we should be investing in battery technologies, proper battery technologies. We should be creating data centers such that we are not using on demand compute, not creating servers that are over-provisioned and under utilized. We should be creating technologies like this as an industry stating that our aim is not to waste the electricity that we're given."
Paul Johnston: But actually we don't do that. We just create technology, and the way that we create technology because effectively it's free, and there's an individual's choice. I have cut down on the amount of travel. I certainly don't fly in Europe, anywhere near as much as I used to. I try to use the train in Europe. Transatlantic travel is very, very difficult as Greta Thunberg has shown. Um and so you usually just have to use a plane for that point, so you've learned to offset.
Paul Johnston: So you learn to use offsetting various offsetting companies, and there's a gold standard for offsetting which you can use. So I think when the when the finger is pointed, I think you just have to turn around and go, "Well, what else are you doing? I'm doing all of these other things. What else are you doing?" And yes, no one person's life is ever going to be completely free of all of these problems, because our countries and our systems of government and our ways of life, our globalization is so full of sunk carbon, for want of a better phrase, that we can't get away from.
Paul Johnston: It's virtually impossible to get away from having a carbon footprint that isn't sustainable for the long term. So unless we change our ways of looking at life and the universe and everything, then I don't think we're going to do anything. There was a talk yesterday that I went to, where someone said, "Effectively, unless governments and a few very big companies change, then the only other way of doing this is to eh is to get the entire world to go to put solar panels on their roofs and to stop driving cars and to stop flying and to effectively all go vegan."
Paul Johnston: That's pretty much what we've got to do, and we've got to do in the next five to 10 years, and I'm pretty sure that's not going to happen. So we've got this problem where we need to fix it, and we don't have the tools to do it. So when someone points the finger, I think we just have to say, "Well, do something but don't do nothing." I think that's the best way of looking at it.
Corey Quinn: Migrating between regions, let alone between different cloud computing providers, is a massive undertaking for almost any-
Paul Johnston: Absolutely.
Corey Quinn: ... environment that is more than trivial to scale. It almost feels like that is far less likely to happen than enough people shaming Amazon into stepping up on the climate front.
Paul Johnston: Yeah. I completely agree with that. I've seen you know I’ve seen data centers that are basically stuck in buildings that are never going to get moved, or at least ... because they'd been there 20 years, and the person who's there is basically ... there's only one person in the company that really knows how it works, and the cost of moving it would be an absolute nightmare but they're in production.
Paul Johnston: I can imagine there are cloud environments now that have been sitting on ... EC2 instances, or moved between EC2 instances and various different things that have been there, what? 10 years now? The people can't move because they don't really understand how they work. And I can imagine all of these scenarios for the non trivial things that we're talking about. We can't just go, "Oh, I'll just move. I'll just do this." It's too simplistic. We've got complex technology, and I think building a movement around this, I think is the best way of going forwards.
Corey Quinn: So what can people listening do?
Paul Johnston: So we’ve got a number of different things, but the ... I think probably the easiest thing is to go and learn, go and have a look at what's out there, go and read some blog posts, go and follow some climate scientists. I think before you do it, and I think this is especially true in America, I think leave your preconceptions at the door. I think we have the same in the UK. I think it's all around the world. If you go and read the IPCC report from last year, which was the special report on global warming to 1.5 degrees and above, I think you'll find that that was signed off by every nation of the world.
Paul Johnston: It's not a political report. It's a report that basically gives you the information. Go and read that, learn something, don't listen to the politicians who are using words that are mired in alternative meanings. Go and figure out what those words are basically hiding, and then go and do the reading for yourself. I've got a Medium post on that I wrote a few days ago actually, which was about stop being a techie and start being a human. I think the tech world has a tendency to hear something like this, realize that there's a problem and then go, "Right. How do I fix this? What's the solution? How can I build a website, an app? And then that will be the thing I need to go and do."
Paul Johnston: And I would just caution against that simply because the problem is so huge and complex that that's not really going to solve anything. You might solve one problem but then there are going to be 100 other problems out there. So I would suggest probably going and sitting and listening to whoever your nearest climate activism groups are. So your 350.orgs, or your ... whoever they are within your group. So with Greenpeace, or Extinction Rebellion, or I guess the Sunrise Movement, and all of these organizations, whoever it is.
Paul Johnston: Just listen. You don't necessarily have to agree with them. Just listen. There are huge number of amazing climate people. There's a climate Twitter that is just full of amazing content. Just go and start to absorb this content, read it, take it on board, and you will start to find that there is an undercurrent of hope that we can do something incredible, but it's an organizational start, at present. We're not at the point of really doing anything. We are gathering, and I think we're in a gathering phase at the moment. But I would just caution against too much doing and not enough thinking and waiting and listening.
Paul Johnston: I know that for techie people that's quite a difficult thing to hear, because we all want to just build something and try and 10-X it and get some funding. But I don't think that's what the world needs right now. I think the world needs people to start spreading messages and talking. I think that's a very different thing to the way that we're used to do in the world now.
Corey Quinn: This is usually where we try to have something uplifting, and oh, here's how it's all going to be okay, and I feel like I'm grasping a little bit to find that narrative.
Paul Johnston: I know. And okay. So let's put it this way. The tech industry is full of people with money, with huge brains, and with a huge amount of love for the world that we live in. We have the opportunity to use those skills and the skill sets of running, building businesses, of taking out ... of building organizations and communication and technology, and taking those skills and putting them to use in the right ways, and in the right places.
Paul Johnston: I think there is the scope for turning is often the industry that people turn around and laugh at and see an awful lot of bad things. The disaster capitalists that effectively that ... A lot of people look at tech and see an awful lot of money and not a lot of goodwill, and I think we've got an opportunity to say actually there's an awful lot of good people in tech. There's an awful lot of good things that could be done. I know that it feels like there's not a lot of hope, but there is an awful lot of good stuff out there.
Paul Johnston: For example, if you've never come across Project Drawdown, so one of the things that we need to do to essentially turn back the clock on climate change is we need to stop emitting carbon. But we don't just need to stop emitting carbon, we need to also be taking carbon out of the atmosphere, so carbon capture and storage. We need to be reducing the amount of carbon that's coming out of various different other places around the world.
Paul Johnston: For example, someone did a plan for doing this and it's called Project Drawdown, and it's got a whole list of things that would help in the climate change movement. Most people would think, "Well, it's basically decarbonizing the grid so it's clean and making everyone vegetarian and vegan," and actually it isn't. There are things on there like educating women in Africa and Asia, and there's things in there around better chemicals in air conditioning units, and there's things around ... It's a completely different set of skills and technologies and understandings.
Paul Johnston: It's also the wind farms, and it's also the batteries, and it's also the things that are obvious. But there are things in there that are tech mind and a set of tech companies could pick up and go, "Actually, I can do something around that. I could build something. I could create something in that space." It's not going to be your venture capital funded, you know 100 billion unicorn company necessarily. But it might be one of the bricks in the wall that helps us do something about climate change in the world.
Paul Johnston: You may get to have a more fulfilling life than helping one of the big companies make you know another 100 million of people playing a game on an app, on a mobile phone that they bought for $1,000 in an Apple Store. We can do better with this industry. I have a huge amount of hope that this industry, it's attracted all the clever people. Maybe all the clever people can go and do something to fix the world we're in now. I have hope.
Corey Quinn: I guess that is the uplifting narrative that we look for in this. I'll include links in the show notes of the article that you wound up writing. The white paper
as well. Where can people go to learn more about you?
Paul Johnston: So I’m on Twitter, @PaulDJohnston
, and on Medium at Paul D. Johnston
. Yeah, just come and find me there, ping me a message if you want to know more. There is also the ClimateAction.tech Slack group, which is well worth getting involved in. There's an awful lot of tech people on there who are trying to figure out how to use their tech skills for this kind of thing, for climate change and for good. So I’ll be on there as well. Come and find me.
Corey Quinn: Excellent. Of course, you are one of the organizers/founders of ServerlessDays.
Paul Johnston: Yeah, I am. If you want to get involved in doing the Serverless thing, which I still enjoy doing, then at Serverlessdays.io
, there's going to be one somewhere near you. I'm pretty sure. So go and have a look, and if you want to become an organizer or get involved, get in touch by the website as well. While we've talked about climate change, Serverless is still a big thing. I've written a lot of Medium posts on that as well, but we haven't talked about that at all.
Corey Quinn: Excellent. Thank you once again for taking time out of your evening to chat with us about this.
Paul Johnston: Thank you.
Corey Quinn: Thank you for that-
Paul Johnston: Appreciated.
Corey Quinn: ... uplifting story.
Paul Johnston: I've really enjoyed it. Thank you for allowing me to speak and to have a bit of time to talk about it.
Corey Quinn: Paul Johnston, interim CTO, and currently working on something stealth mode, and not a crank. I'm Corey Quinn. This is Screaming in the Cloud.
Speaker 1: This has been this week's episode of Screaming in the Cloud. You can also find more Corey at screaminginthecloud.com, or wherever fine snark is sold.
Speaker 4: This has been a HumblePod Production. Stay humble.