Understanding AI-Powered Simulated Worlds with Christopher Entwistle

Christopher Entwistle is an Executive Producer at Improbable, a cloud-based platform for game developers that revs the creation of AI-powered simulated worlds. In this slogging AMA, we chat about the Improbable, Chris' role at Improbable, the future of the metaverse and how to make money in it.

This Slogging thread by Mónica Freitas, Chris Entwistle, Cris Silva, Jack Boreham, Sara Pinto, Limarc Ambalina and Abeer occurred in slogging's official #amas channel, and has been edited for readability.

Hey @channel, please join me in welcoming our next AMA guest, Chris Entwistle. Chris is an Executive Producer at Improbable. Today, he joins us to discuss how cloud-based platforms are essential for game developers to accelerate the creation of AI-powered simulated worlds and big-picture topics like making money in the metaverse. Chris is happy to answer any questions you have on Improbable, what sets Improbable apart and how its technology is solving key challenges facing the gaming industry.

Please feel free to ask Chris anything about:

Hi Chris Entwistle! It's great to have you here! Could you start by telling us a bit about yourself, your role in Improbable, and your background?

Thanks, Monica! And thanks to Hackermoon for having me - this AMA is a new experience and I'm really looking forward to it. 🙂

I'm an Executive Producer at Improbable working on Morpheus-related products. Morpheus is one of our metaverse-related products that allows 10,000+ players to be in a single server instance in real-time. If anyone is interested in what that looks like, here's me calling out BagginsTV last year in a 4,000 player event in Scavengers (https://www.youtube.com/watch?v=kzxTcHqqLoI). Calling out a streamer live was one of the highlights of my year.

My background has always been in games. I didn't go to college or university - I actually got a job at McDonalds working minimum wage while I tried to figure out what I wanted to do in life. A job opportunity came up as a games tester for SEGA on the game Metropolis Street Racer back in 2000, and I fell in love with the industry in my first week there.

Since then I moved through roles as a systems designer, level designer, producer, programmer, and even co-owned a VR business at one point. I've been super lucky to work with some of the most talented and creative people in the world, and along the way, I like to think I've absorbed a lot.

So yeah - looking forward to answering any questions this community has. I'm an open book and would love to talk about the games industry, the metaverse, Improbable, or anything involving my experiences in the world of technology.

Hi Chris Entwistle! The metaverse is still a bit of an unclear vision. What do you think will be the future of the metaverse?

Cris Silva - 100% agree, in my opinion, the metaverse is a bit like that famous Bill Gates interview on Letterman where he's talking about "the internet".

The final form of the metaverse isn't something people can visualize in a concrete way, because it's a lot of interconnected things and we don't know what the interaction of those things will produce.

However, what we can do is look at some of the baseline requirements. One of the better articles I've read about the metaverse is Matthew Ball's. It describes the core enablers for what the metaverse is likely to need:

If you look at the "Networking" section you can see why Improbable is so interested in the emerging potential of a metaverse layer - we've spent nearly a decade working on the problem of making complex simulations easier to manage and usable by developers. It's wonderful to us that people are starting to see the potential in what this enables.

But to be more direct about what the future of the metaverse will be: my personal opinion is that we'll see an upswell in earning a living online.

To be clear, I don't see this as most people spending their time trading virtual goods, or grinding gold like they do in MMOs. Both of those things can arguably be done easier and cheaper using computers - which means they'll never become a job for masses of people.

I think we'll see entire new services appear instead. Just think about the rise of the Twitch streamer - this was a role that didn't exist 15-years ago. It's hard to make a computer do that role because it's a personal, human service that they are providing.

I believe the metaverse will provide more opportunities along these lines, giving birth to a whole suite of ways for people to earn a living doing things they enjoy.

Chris Entwistle, what a great journey! What would you say were the biggest challenges entering the gaming universe? And how did Improbable come into your life?

Mónica Freitas - Oof, that is a big question.

Well, talking specifically about right now I will say that the games industry hiring scene is hot. It's never been a better time to get a role within the games industry as demand is outstripping content enormously.

Here's a fact: the games industry is larger than the music and film industries combined, and has been so for many years. Every studio I know is hiring - the players want more and more content to play, and that trend is only getting stronger.

As to the biggest challenges: I'd say probably a few things:

1> You really have to want to be in the games industry. It's a highly iterative, collaborative, creative process - and it comes with a lot of uncertainty and risk. Only 1/10 games ever succeeds, so you need to enter every project knowing there's a 10% chance you might not make it. It's not for everyone, but the rewards of seeing players love your project? Absolute gold. Personally, I live for that.

2> Learning the tools is important. Someone wanting to be in the games industry should really have at least a passing familiarity with the tools used, e.g.: Unreal, Unity, Cryengine, etc... The architecture and the design patterns we use aren't always intuitive, so having even a small amount of experience helps a lot.

RE: How I ended up at Improbable - honestly, it was on the back of a failed project I'd just completed.

I spent 4.5 years at BioWare working on Dragon Age Inquisition which was a great success for that studio. I needed a break so I split from BioWare and started a VR company focused on creating games about open-world exploration. There were only three of us and we were learning C# programming as we went, and as you can imagine... it was tough.

But we learned a huge amount, from creating basic avatar systems, to packaging analytics data for VR players to AWS, to figuring out how to actually do asynchronous loading systems in VR that don't kick you out to the Steam main menu and jolt your brain.

Improbable reached out as we'd finished a project that didn't take off, and they were super aligned with what we were trying to achieve. After conversations with the team, I realized we wanted a lot of the same things, so I agreed to join and have spent 3-years chasing the same goals but with an amazing team of talent to help.

Chris Entwistle, great to have you here. So how does improbable seek to enable the Metaverse? 😄

Jack Boreham - well, we've got a few advantages that we're hoping can help enable the metaverse.

For a start, we can have 10,000 unique players in real-time in a single Unreal server. On top of that we have demonstrated the ability to add:

A lot of game developers begin by actively trying to avoid situations like the one we describe above, e.g.:

We started from the position of: "Ok, but what if we HAD to allow all those things - how would we architect a system that could handle it?" And it was tough. It was like inverting your brain and unlearning all the things you knew you weren't allowed to do.

But we got it working and it opened up totally new experiences - like having meet and greet events with the developers and their entire CCU for that game (we announced game updates for Scavengers inside the game at one point).

That's one way we think we can help with enabling the metaverse - we've made headway on solving one of the major problems, allowing people to be together in real-time in numbers and order of magnitude bigger than other systems out there.

We have a saying at work: The Metaverse is better with a crowd.

And you know what? From all our testing and live events so far, it really, really is.

Chris Entwistle Thank you for your response. Those are excellent points. I do struggle a bit seeing in coming to fruition as it demands such a big structure to become a reality... Plus, the security aspect throws me a bit off, as we all know the issues Facebook had with data protection. When it comes to making money on metaverse, how do you think that will go about?

Cris Silva - This is the literal billion-dollar question. I'm going to flip it though as I think it's easier to work through that way.

Making money in the metaverse - what value will the metaverse provide to users, and therefore what will they be willing to spend money on?

For us to be able to generate anything of value it needs to fulfill a few criteria:

To truly make money in the metaverse, whatever it's offering needs to fill the criteria above - so how about an example...

Online Concerts w/ Persistence
Think about a real-world concert - why do people go? They could just stream the latest album on Spotify and arguably get a crisper, clearer version of the music on-demand.

The answer: the social experience. When people go to concerts they want to hear the roar of the crowd, they want to hear live music and shoutouts with their friends ("do you remember when he called us up on stage that time!"), and they essentially want to generate shared memories.

Real-world concerts are awesome and I love them - the swell of the crowd, the vibe of the room, the overpriced drinks - it's all great. But it has some drawbacks:

Virtual concerts will never replace physical ones, but they could exist alongside quite easily - and they definitely bring value:

Remember - there is an entire generation of kids who've grown up playing Minecraft and Fortnite, where meeting their friends online in Discord isn't just normal, it's expected. If you can find what they value and give them easy access to it they will spend money.

Now add persistence to that - you get a custom T-shirt for attending a gig, or a special "fan club" ticket access to the secret server - and you're building long-term value on top.

Chris Entwistle, that's some great news for game developers and artists. I had no idea how big the gaming industry was.

Chris Entwistle so, your adventure through VR gave you many different notions on the future of gaming. Are trends in gaming that you're noticing?
Moreover, VR is still in an "early" stage. We have many theories about it mixing with AR, but as someone outside of the gaming area, I'm not even sure what's possible and feasible to accomplish in the next ten years. How do you think it will evolve?

Mónica Freitas - I'll start with trends in gaming.

A big one we're starting to see is the walls coming down between platform holders and also different IP holders.

Dead by Daylight brings together monsters and villains from multiple high-profile IPs. Fortnite has Star Wars, Ariana Grande, Travis Scott, and the Avengers in the same package. Nintendo, Microsoft, and PlayStation are allowing their players to connect with each other through cross-play.

I think the simple thing here is: the whole is greater than the parts. We're seeing big businesses recognize that keeping everything walled off from each other actually hurts the industry on aggregate.

By allowing brands to exist alongside each other and players to connect together, the whole system itself becomes stronger. This was pretty much unthinkable in 2010, whereas now it's becoming the norm.

For VR - I think the biggest problems we had when we had a company pursuing the technology were largely physical. If you could get someone to put the headset on they almost universally agreed it was magical - but getting them to that point was tough because:

This is where I believe VR has the biggest chance to grow. If VR companies can make it easier to put the headset on (and AR glasses are a great step towards that which is why I understand the Magic Leap hype) then it brings 3D spatial virtual worlds so much closer to mass-market adoption.

Make it less socially weird, make it less uncomfortable to wear, and reduce the space requirements.

Hi, Chris Entwistle, glad to have you with us today! So, I'm curious, how can AI improve gaming and power simulated worlds?

Sara Pinto - AI is a very, very interesting topic. It can be used in a multitude of ways to improve things in the gaming space. Here are a few areas I know that gaming companies are looking at:

Modern-day animation https://i.imgur.com/JPm2rQX.png. The image in the link to the left of this is actually a fairly simple example of an anim-graph authored by a developer - if you want to get to Naughty Dog levels of animation fidelity you're talking about anticipating thousands of potential situations to pick the right animation to play. With AI a computer can read what your in-game intent is and predict the animations it needs, load them into memory, and play them if it gets them right. This increases performance, fidelity, and responsiveness all at the same time.

Just think about how often people interact with social media, and how many data points Twitter, Facebook, etc... collect from you. If you think that's rich data, it's nothing compared to how much data you're creating when playing a game. What items do you interact with, what playstyle do you adopt, how do you interact with players, what skill level are you at with each challenge, etc... This data is a treasure trove for designers, and with AI there is the potential to look at what you enjoy and serve you that content, leading to better experiences and higher retention.

Last, but not least - multiplayer games are intensely linked to network performance compared to things like Netflix, social media, etc... 200ms of lag can ruin an experience. AI can be used to help shape how network optimization is performed in order to use the least amount of data for the maximum effect. It's an area Improbable has a lot of experience in.

Chris Entwistle, so cooperation is in, which is good. But what do you think of big merges in gaming? For example, we've seen a lot of significant acquisitions in the gaming area, like Activision and Bungie. Do you feel that this smothers smaller gaming companies?

Mónica Freitas - RE: mergers in the games industry. Honestly, this is an area I'm genuinely not an expert in, so think of this answer as more of a gut instinct guess than one based on an in-depth analysis of the market.

I've been in games since the year 2000. In that time I remember the industry going through periods where the major publishers would snap up indie developers and take their technology in-house. You should have seen the reaction when EA bought Criterion and that took RenderWare off the market in ~2004; a LOT of PS2 games were built on that software and for those developers, the sky was falling.

One thing I will say though - these days indie developers have much more ability to design, develop, and distribute games without the need of a publisher. Steam, EGS, Kickstarter, social media platforms, memes - there are a lot more tools now (and a lot more gamers) which allows for much more niche games to be successful.

All of which is to say: I don't think mergers and acquisitions in the gaming space is anything to be alarmed about - they've happened in a cycle of every ~8-years since I've been developing games, and the games market has only got better in that time.

Chris Entwistle, you're right, virtual concerts have grown since the pandemic, and it's probably a reality we'll live with from now on.
I wonder how differently the metaverse will impact brands and marketing. I'm sure it will affect how we do business somehow.

Cris Silva - RE: Marketing in the metaverse - I think marketing companies/departments are always looking for a few things:

How many people can they reach by taking part in your metaverse project? At the end of the day, marketeers need humans to see the product they are selling (or at least have someone see it so they can tell a friend). If a metaverse is successful on paper but doesn't actually have any players using it, it's not going to be a great pitch for brands to bring their content in unless they're looking to write blogs/tweets about it (which will have the reach they want).

The next thing we care about is the quality of impression that you can give in the metaverse. A virtual billboard is cool, but a moving billboard is more effective. Better still is an item in the metaverse that actually has a positive function, e.g.: your Nike shoes actually make you run faster.

A big one right now is "how shareable is your metaverse"? Does it generate memeable content, in which case do brands have the opportunity to be part of something that happens in the moment? Worlds that are alive with content generation will be rich with opportunities like this.

As stated above when talking about AI, games have a lot more data points we can collect versus social media, and we're only in the beginnings of trying to use that data.

Marketing in the metaverse is an extension of what marketeers already look for with buses, social media, billboard, and print media - opportunities to reach an engaged target audience.

Hey, Chris Entwistle, thanks for joining us! I'm really big into VR and am interested: what VR business did you co-own?

Limarc Ambalina - I co-owned a company called PrimeVR that worked on two projects:

The other co-owners also leaned into some blockchain projects in 2017 that were interesting (including and NFT type project on the Bitcoin blockchain). It was fascinating to see that scene come alive with developers at the time.

Chris Entwistle thank you for your detailed answer. You talked about unreal, is that unreal engine 5?

Jack Boreham - We work in both Unreal 4 and Unreal 5. Thankfully Epic made it easy to work in both, if you understood either.

Unreal 5 is pretty incredible as a piece of technology too btw. I'm somewhat blown away that it's out there for free - the amount of software developers get to use in game creation now is incredible.

Chris Entwistle, thank you for the detailed answer! The fact that AI can improve all these, it's amazing! AI is complex software. Investing in it in these areas could bring any cons?

Sara Pinto - Very much so - AI is an emerging technology in the video games space (at least, the machine learning type). You need to do a lot of upfront work to integrate AI into your systems:

By removing the human from the loop you're essentially giving control over to the machine without fully understanding what's in the black box, and that requires a different set of training and development processes from the norm.

If you have the time and budget to do it though, it can provide spectacular results and long-term value for the project.

Chris Entwistle, it’s great to have you here. Morpheus sounds like a game-changing (no pun intended) product. Is there a video game concept that you think can take full advantage of having 10,000+ players in a single server instance? (e.g. perhaps a shooter game that simulates a war using only players and no NPCs)

Hi Abeer! Let me take you on a journey of discovery...

One of the first things we wanted to try was actually a bit of a dream job. We wanted to create a massive pitched battle like Lord of the Rings with an army of human players facing off against hordes of AI - and we did!

If you skip you can see a short clip of me running around like a giant blue idiot trying to organize an army with a baseball bat.

But, the thing is - the human players decimated the AI. Like, they chewed through something like 18,000+ in about 2-minutes. We tried dialing up the health and attack power in later events, and the humans still nailed them. We realized that while this experience is super interesting and has some huge potential, it also incurs a very large amount of developer work upfront. It was a hard thing to balance because you need real players to balance it.

So we moved on to some other ideas like this one:

This is essentially 10,000 player Fall Guys. This was like American Ninja Warriors but with everyone on the course at once, and it was a lot easier to create and balance as a developer. We even added a pink/blue team-based mode to let players win or lose as a group. The simple platforming lent itself well to having crowds of players.

Other things that worked well: celebrity meet and greets. Because of the one-to-many voice communication, we could bring in developers from the Scavengers team to talk directly to the players, and this generated a huge amount of positive sentiment. You were in the build, with the devs, and you could talk to them and engage. We extended this to Kpop star AleXa and did a global fan party with her live with ~1,200 of her fans.

And finally, we had a behind closed doors game mode which I can't talk too much about - but it was a streamer vs streamer game that allowed them to compete and bring their entire audience in to fight at their side. Just imagine Shroud vs Ninja, each with their fans hip-to-hip with their idols in the game. That's something we still believe in, so watch this space. 😉

In short, though - we don't think there's one video game concept that this technology enables, we think there's a whole suite of new concepts that we can now look into. I'd be super interested to hear your thoughts on what you'd do because it's always fascinating to think about what experiences people want.

Chris Entwistle Thank you for the answer. I see that the video game industry is under quite the changes. How can one profit from this? Are able to make money in the Metaverse?

Sara Pinto - I think my answer to Cris Silva's question further up the thread is about the best I can do RE: making money in the metaverse 😄

I think there are many ways to make money, but to do so we need to establish what the users value and find ways to serve it to them. If we can give unique, genuine experiences to people and allow them to carry that value around with them to other experiences, there will be opportunities for developers to make money.

Chris Entwistle, thank you for your insightful answers! It's been so great to have you on, but unfortunately, we will have to wrap up this AMA. Do you have any final thoughts for our readers or something you'd like to promote?

Mónica Freitas - Thank you so much for having me; these were great questions and it was wonderful to be able to talk a little about what we've been doing here at Improbable.

My only final note would be this: there's a lot of hype about the metaverse, but that also comes with a lot of cynicism and negativity. This is an emerging market and there are going to be some amazing wins and some complete failures - but it's probably going to be years until a true metaverse emerges.

In that time I'd advise people to get first-hand knowledge of what's happening - go download a demo, or read a whitepaper, or join a community - it's a vibrant scene and the only way we can figure out the whole is with people participating and making their voices heard. 🙂

Oh, er... actually one more final note.

People seem to conflate "The Metaverse" with "VR" and I think that's a mistake. When the metaverse occurs it'll be accessible from your PC, your console, your phone, from AR and from VR. You'll get updates on your watch.

We're a connected society now and the metaverse is a meta-layer across it all - so the metaverse will have many access points, not just a VR kit.

Ok, I promise I'm done now. 🙂