Our Goals for 'AI and Games' in 2025 and Beyond! | 13/08/25
Time for a 2025 Subscriber Drive - with 20% off paid membership!
The AI and Games Newsletter brings concise and informative discussion on artificial intelligence for video games each and every week. Plus summarising all of our content released across various channels, like our YouTube videos and in-person events like the AI and Games Conference.
You can subscribe to and support AI and Games, with weekly editions appearing in your inbox. If you'd like to work with us on your own games projects, please check out our consulting services. To sponsor, please visit the dedicated sponsorship page.
Greetings and welcome to another issue of the
newsletter. This week I’m dusting off an idea that we’ve not ran across AI and Games for the past couple of years (at least in its original YouTube-only format) and bringing it to our written offerings. A subscriber drive as we push to make AI and Games bigger and better.So to do that, something on an introspective issue in which we provide:
An inside look into the ins-and-outs of running this here project.
How we’re getting on as a creator-owned and reader-funded website.
An insight into broader business and how things sit.
Our plans for 2025 and beyond for what we’re doing in this newsletter.
How you can help support us, and what we plan to do with your help.
So in a reversal of how things usually run around here, the big news story this week is at the top, and we’ll dig into other announcements and the news at the very end.
Follow AI and Games on: BlueSky | YouTube | LinkedIn | TikTok
Support the AI and Games 2025 Subscriber Drive!
Today this issue launches around 18 months since the newsletter started (just shy of 70 releases on the website, with the first few on LinkedIn), and we’ve been excited to see a lot of growth surrounding AI and Games in the past year. And critically we want to do more of those things, and a lot of that comes down to you, our wonderful readers, and your support of our endeavours.
As is highlighted at the top of every issue, we’re a reader-funded endeavour: this newsletter happens thanks to you and your support, giving us some leverage to allocate time towards it. To that end, we’d love to do more, but we need your help to do that.
So we’re publishing this issue to give everyone both an inside look at how our business operates, but also our broader goals as part of a subscriber drive. The more support you can provide us, then the more we can do here on AI and Games! We’re currently running a 20% discount for the next month for annual subscriptions, and we’ll be sharing our plans in this issue of what we hope to achieve with your continued support!
AI and Games in 2025
So I figured let’s start by discussing things are with
right now. Over the past 18 months we’ve expanded out into more content creation, and critically this here newsletter. A big part of this is that we recognised people are looking for meaningful and insightful discourse on artificial intelligence in video games. At a time where it’s becoming one of the most discussed topics of the period, there is very little meaningful content out there that actually gets into the meat of the subject matter while equally being accessible to a broader audience.AI and Games started out on YouTube in 2014, but we recognised that our video essays like the one you see above - which we’re very proud of - don’t allow us to react or pivot quickly to things in the news. As you can see across our history of YouTube content, we’re focussed on examples of good practice within the industry, interesting research projects, and how people have solved technical and design challenges. The YouTube channel started as a means for (aspiring) game developers and scholars to learn more about how their peers solve similar problems, and now it’s built a strong reputation in industry circles despite never becoming a ‘big’ channel in the grand scheme of YouTube’s gaming ecosystem. An outcome that frankly I never imagined but am very grateful for.
But beyond the editorial focus, making high-quality videos like ours takes time, and I wasn’t for jeopardising the quality of the writing, editing, and production value of our videos in pursuit of ‘reactionary’ or periodic content. I mean the recent Alien: Isolation retrospective was a project we worked on throughout the course of a full year.
Plus there’s an additional concern: not everyone wants to watch YouTube videos, and so it felt appropriate that we try finding another route to engage with this demographic.
To that end, we launched the AI and Games newsletter in early 2024, and I’m really happy with how it has gone down thus far. It’s given us space to react to events happening in the industry, and tackle a whole bunch of topics to help people understand what’s really going on in AI for games. Crucially, it works in a way that our YouTube episodes - which appear as case studies on the site - are not a good fit for.
Just this year alone we have:
Dug into the ongoing legal consultation on AI for the UK creative industries.
Cut through the misreporting of Microsoft’s Muse Project
Explored the consumer backlash developers are facing over generative AI adoption.
Highlighted the real implications of IGN and Eurogamer’s works appearing in a lawsuit against OpenAI.
Cut through the noise on GDC’s ‘state of the industry’ report, and how it misrepresents AI adoption.
Discussed the collective failure on assessing AI adoption is the reason why platforms are being flooded with slop.
Plus recaps of attending events such as GDC 2025, the summer school on AI and games, and researcher’s digest issues on Google SIMA, Microsoft’s WHAM and the BFI’s screen sector report.
And while I’m really pleased with this, and we need to continue to push to do better. In truth, we’re largely at capacity of what we can do at this time. The newsletters and YouTube work take up a good chunk of our time, and while they’re very much the heart of everything we do, they’re also the least lucrative - a point I’ll return to in a second.
I Hear You Like Statistics?
But first I wanted to share some numbers about how people like your lovely selves have embraced what we’re doing. Here’s some quick stats on where things are with the newsletter right now:
Running at just over 3,300 subscribers to the site as of August 2025
A 135% increase since August 2024.
Averages around 2,200 reads per issue from June through August 2025.
An increase of traffic by 115% from the same period in from 2024.
It’s been really satisfying and rewarding to see the broader impact that the newsletter is having. We’re engaging with an audience that seldom, if ever, would watch the YouTube episodes. This isn’t hyperbole, we have been following the network traffic. Only around 8% of our traffic to this site in the past year comes from YouTube (where we also post links to the newsletter). Equally, while our external referrers are a small amount of our total YouTube traffic (9%), our own website is only our 6th most likely referrer to our YouTube channel with around 10%.
What we have seen however, is that social media and web searches have been our most successful traffic mechanisms. People are finding out about our work and signing up for it - I hate to say it but LinkedIn of all places is one of our biggest social network referrers.
For me personally, one of the funniest parts of this is that I now have conversations at industry events where people recognise me as the author of the newsletter and not the YouTube channel. This is amusing because for years, outside of the game AI community, this was largely how I was known within the sector. The Venn diagram between our newsletter and video audiences does not overlap as significantly as you might think, which is quite a positive for me, given our 220,000+ YouTube subscriber base isn’t always game industry professionals - and while we love them dearly, they’re not the target audience for the newsletter. In fact right now 32% of our readership overlaps with the
, which is quite fitting given a) and I are good chums, but also b) we’re both trying to tackle the same issue of delivering valuable insights to the industry.The Business Reality
Okay so, to summarise everything thus far:
Newsletter getting good increase in traffic.
Feedback thus far generally very positive (anecdotal but hey, facts)
Newsletter reaching our target demographic (professionals interested in AI and game development).
Now it’s worth stressing that all of this sits atop the work that we already do in the games industry, plus of course there is Goal State, and the AI and Games Conference - and in truth it speaks to how we find ways to make this a sustainable business such that we can create content.
As mentioned, I’ve been running AI and Games in some form or another since 2014, and the ‘content’ side of things - i.e. the YouTube videos and this newsletter - has never been financially sustainable. It’s why even at the height of our YouTube popularity around 2019-2022 where we getting millions of views a year I never contemplated becoming a full-time content creator. I recognise that the work we do has value, but also it exists within a niche that means we wouldn’t be able to retain our editorial integrity and scale it to be financially successful. Plus it’s at the whim of things like YouTube’s recommendation algorithms which don’t really value long-form technical subject matter.
What keeps our company solvent, is that the vast majority of our income (80% in the 2023/2024FY) comes from the work we do in industry: consulting with games studios, running professional training workshops, and even some of my speaking engagements around the world. And it’s worth stressing this isn’t some master plan that I had back in 2019 when the company was formed, this all happened by accident as games companies starting reaching out asking me to come and do work with them, and we realised this was a more realistic means to build a business and have impact on the games industry rather than be another hetcis white male posting opinions about game development on social media every other day - I mean, we have enough of those. Surely?
Goal State and the Conference
Meanwhile in 2024/25 FY we’ve made the pivot to building product, given I recognised the market volatility is affecting our primary revenue streams. The games industry is a mess right now, and we’re working harder to make the same amount of money year-on-year as budgets tighten (contrary to all the noise about AI investment).
Hence we came up with the idea that is now Goal State, and the crowdfunding has given us some financial buffer to get it done and then get it out there - and in the meantime we need to keep making money so we’re still financially sustainable after it’s done.
Meanwhile the AI and Games Conference is still very much in its infancy, and we make little money from it. This is because:
It’s a separate company (Game AI Events CIC), and while it uses our brand, I’m only one of the five founders.
For context, we license the AI and Games brand to the CIC for free.
It’s a community interest company, and in turn a community endeavour - so not only am I not seeking to nickel and dime it at the expense of putting on a good show, being a CIC means restrictions on how our profits are spent.
It’s still so early in its development that we’re still trying to establish a financial footing for it to scale up as a successful enterprise in its own right.
I have it on good authority that running a conference and having it sell out and turn a profit in its first year is not common. We turned a tiny profit in 2024, which was just enough money to buy us all dinner, and get started on planning 2025.
This is a long way of saying that the thing that we’re largely known for, as content creators, acts as a loss leader: it keeps us in the conversation, given we’re helping to craft it, and then use that to get eyes on our primary business.
Competing for Time
So this brings me to the last part of the more formal business discussion bit: it’s all about our available time. We’ve scaled up a little this past year, to the point I get to saw ‘we’ a lot in this issue. It’s not just me doing this full-time now, but even then we’re at our limit of what we can do.
It’s a bit of a Catch-22 really: we’re able to maintain a roughly weekly newsletter cadence, and a roughly monthly video cadence, but still we need to focus on our client work given that’s what finances the content creation in a sustainable way.
On that note, there will not be a newsletter next week because I am at Gamescom in Cologne followed by IEEE CoG in Lisbon, and I have a bunch of client work to wrap, meetings to attend this week before flying out, plus two presentations to finish writing. My time is often quite fractured, and as such the content creation is often affected because our primary business is where the money is.
And so for us to grow the content creation side, we need to try and boost the revenue it generates. We have on occasion ran sponsorship on our newsletter as you’ll have seen this past year, and those have been great partnerships and put some cash in the bank. But in truth we reject over 90% of sponsorship offers that come our way. They’re typically of low quality, not a good fit for our more discerning audience, or are simply not worth the money given the potential reputational hit we might receive.
All that said, if you think your business is a good fit for sponsoring us, get in touch!
2025 Subscriber Drive!
Okay, enough of the serious business stuff, let’s talk about what we’re trying to do here! In short: we want to grow what we’re doing here on AI and Games, including:
More reporting on relevant topics.
Editorials and insights into industry goings on.
Given you, the reader, more opportunities to influence what we produce.
Bring in other - fun and interesting - voices to discuss issues and topics relevant to our audience.
Ultimately do more stuff that is of interest and relevance to you, without relying on me - i.e. the creative bottleneck.
At the time of writing we’re sitting at 3327 subscribers - people who sign up for regular updates sent to their inbox - and 54 paying subscribers at around £5 ($7) a month. You’ll note the disparity: a conversion rate of 1.6% of our readership.
I will clarify that we historically have received fan-funding courtesy of our Patreon page, where we have 93 paying subscribers - 4.5% if you assume our patrons read the newsletter. Though we raise more funds from our 54 site subscribers given most of our Patrons are on the lowest £2 ($2.75) tier.
Goals, Goals, Goals
Okay, let’s get into it, we want to achieve two things:
Increase our reach and audience.
Encourage our wonderful readers to support our growth!
To do that, we figured why not take page out of other independent sites and share where we’d like to get to, what you get in return, and we’ll hold ourselves to what we’ve discussed here. So let’s get into it:
Subscriber Goals
So let’s go through the first phase of goals, regular subscribers. The more people that engage with our material month-in, month-out, then this will help us grow, and we will introduce new content as our audience follows us:
Nothing too drastic here, but right now the bigger the audience becomes, we want to give you more of a say in the types of content you’d like to see on the site. After all the more readers we gather it’s often a challenge to then discern what your interests are. So hopefully we have get more of an insight from you.
Though equally, get down in the comments of you have thoughts on what you’d like to see from us that we haven’t thought of!
Paid Subscriber Goals
And of course for all of our wonderful readers who support us, the key thing is we’re looking to expand into writing more material while also expanding who is making it. Both in terms of production support, but also having more writers come in on the publication. This is important to me personally, because I’d love for us to use this platform as a means to support new voices in the space such that it’s not just my perspectives you’re hearing week-in, week-out. I’m conscious that I’m another DOWD (dusty old white dude) that is something of a prominent voice in my corner of the industry, and I’d like to use that platform to do good!
So let’s look at those goals…
54 - We are currently here
100 - Launch a new AI and Games podcast summarising our written work - with some additional editorial - in audio format.
150 - Start running periodic pieces from freelancers helping us to expand into more investigative journalism.
Even better, your support is what helps us pay them!
200 - We’ll expand our ‘digest’ series to have two monthly issues on both industry and research developments.
300 - Regularly report on events across the sector (that we can’t afford to go to right now) in both academic and industry contexts.
500 - Expand case study content to twice a month.
These are the more immediate things we can do and have had conversations internally about the logistics for, and are confident we can deliver on.
In truth we have even more ambitious goals, but a lot of that is down to the kind of support we can muster. Though I will leave some of the more ambitious and ridiculous ones right here.
1750 - What is needed to allow us to run content creation full-time and not be dependant on other sources of income (based on current staffing/overheads).
5000- We will hire someone - probably a Scottish someone - whose sole job is tour industry events around the world and call out anyone talking shit about AI in the games industry - with fervour.
Benefits of Supporting AI and Games
I mean, outside of the warm and fuzzies you get from being an awesome person?
Well there’s a bunch of reasons to support already, and we get into it right here:
Access to the supporter-only area of the Discord server, where we have our chats, plus links to works in development are shared there in advance. So if you enjoy the YouTube content, you typically get to see it there a few days before it goes live.
Every month we run the sponsor newsletter, where let our supporters get an insight into everything we’re doing, including YouTube content, newsletter topics, the conference, and everything else in between.
Our digest issues of the newsletter go into detail on research papers and industry topics as we use our expertise to help make these topics more understandable.
Plus our new Producer tier here on the website echoes that of the original Producer tier we had on our Patreon page, with this I’ll be kicking off our monthly production meetings once again. Where our producers can meet with me on Discord and have a broader conversation about what we’re up to and help contribute to the programme.
Thank You for Supporting AI and Games
We still have a long way to go, and we have some really exciting topics across both the newsletter and the case studies that we already have in development, and will be coming to you in the next couple of months:
Interviews with investors on the state of AI and games, from both sides!
A deep-dive into the making of an LLM that plays Counter-Strike!
A sit down with developers behind horror game Amnesia: The Bunker
We get the truth out of Meaning Machine, creators of the AI-native murder mystery game Dead Meat.
A chat with developers at Massive Entertainment about their ground-breaking AI tool that helps them predict game engine performance on consoles.
All that and a bunch of stuff not ready to share yet. Plenty more coming your way!
Thank you all so much for your continued support and readership, and I hope you’ll join us in trying to make
bigger and better in the coming months.I’ll be making a habit of returning to this drive every year, and let you know where we are in terms of audience and overall growth, and fingers crossed next year we have some good news to share!
Alrighty, I should go and write the actual newsletter now…
Announcements
Now that we got the big conversation out of the way.
A Hype-Free Assessment of AI in the 2025 Games Industry - Live at Devcom!
My Devcom talk is now scheduled for Monday, August 18, 2025 from 5:45pm over at Stage 15, Confex Level 2. I’ll concede I have no idea what the last part means as it’s my first time both attending and presenting at Devcom, but a quick Google suggests it’s a room with maybe a 200 person capacity?
So yeah, this will be an hour-long presentation digging into the state of all things AI in the games industry. An update on many of my presentations from 2024 in which I talk about more recent events both in terms of AI technology as well as the games industry - some of which discussed in this here newsletter.
I’m still putting the finishing touches to the presentation now, and trying to figure just how much I can present in one hour without it my resorting to speaking at my natural cadence - y’know, ‘Scottish’ - meaning people will struggle to keep up.
Meanwhile, if you’re at Devcom/Gamescom and want to meet, hit the button below, I’m free for meetings until end of Thursday.
Check out the Reality++ Jam 2025
We had a message come in from the team at Plus2Studios who are a small indie team developing indies games and their plans to host the Reality++ Jam: a game jam for teams of up to five people to build a VR game in one week.
We have no involvement in this, but having ran our own game jams over the years, we are happy to try and signal boost for other organisers. You can find out more about the event at the video above, or at the main jam page on itch.io.
AI (and Games) in the News
Alrighty, time to start bringing this weird upside-down, inside-out issue to a close by reporting on the news - the thing we usually stick in the middle!
Riot Faces Fan Backlash From AI-Generated Trailer (in China)
A story that we missed last week saw League of Legends creators Riot Games received criticism from fans after releasing an AI-generated trailer for Wild Rift - the mobile version of the popular 5v5 MOBA - that was shared on the game’s official page on Weibo: one of the most popular social media platforms in China.
As reported by IGN, fans took to the Wild Rift subreddit to express their concern over the quality of the output, the lack of consistency with established characters, and the general unease of AI-generated content and its impact on the development team. The trailer has since been pulled down with Riot publishing a statement stating the “creator-made” video “did not hit the mark”, and doubled down on the studios commitment to maintaining quality in their outputs.
Microsoft Absorbs GitHub into AI Engineering
Arguably the big story of the week: GitHub CEO Thomas Dohmke resigned from the organisation on August 11th, and with his departure Microsoft is making the decision to absorb it into their AI division - an idea that sounds nuts on paper, but also… yeah.
For anyone not familiar with GitHub, it’s a platform for storing and sharing code, and is commonly used for sharing open source software. It started out in 2007 as a means to share projects that relied on the ‘Git’ version control system.
GitHub was bought by Microsoft way back in 2018, but it still largely operated as a separate entity given it had Dohmke as the leader of the organisation. With Dohmke stepping down, the decision has been made - per a memo reported on by Tom Warren at The Verge - to be overseen exclusively by Microsoft’s CoreAI engineering team.
While of course we largely think of GitHub as a site for the distribution of open-source software, Microsoft have largely embraced it as training data for their Copilot code completion tool, not to mention that Microsoft now estimate that around 30% of the code written in their organisation is AI generated.
Sure would explain a bunch of issues I’ve had with Windows 11 updates of late…
Generative AI Fans Hold Funeral for Anthropic’s Retired AI Model
Well, it’s not the headline I expected to write this week, but I wanted to highlight it given there’s something of a story here. As reported by Wired a group of users of Claude’s ‘Sonnet’ LLM attended a ‘funeral’ in the SoMa district of San Francisco as the model has now been retired by creators Anthropic.
Tiktok failed to load.Enable 3rd party cookies or use another browser
Being the home of art exhibitions and just up the road from Silicon Valley, SoMa seemed like the perfect fit for hosting something as ridiculous as a funeral for an AI model, and it does seem less like a serious endeavour and more a cringe-inducing art experience, with LLMs being used to write eulogies and other bizarre applications.
But I raise it purely because it speaks to a broader issue with the most popular AI models, in that while US firms are now moving into the realms of open source releases, many models are still a risk to adopt given they only continue to exist at the whim of the original creator.
OpenAI amazingly pulled off both of these behaviours in mere days. First releasing open source models - gpt-oss-120b and gpt-oss-20b - for the first time since GPT2 in 2019. Then days later with the release of GPT5 they’ve now gated off use of older models via ChatGPT - and most paid subscription tiers as well. The OpenAI hand giveth, and taketh away, with a lot of PR bluster in between.
But anyway, speaking of Anthropic…
Anthropic Pleads Courts to Reconsider Class Action Lawsuit Against Them
So one of the ongoing legal cases that has been worth watching - among many - is a class action lawsuit raised against Anthropic. Submitted by three authors - Andrea Bartz, Charles Graeber and Kirk Wallace Johnson - in Californian federal court in August of 2024, it states that the company “seeks to profit from strip-mining the human expression and ingenuity” behind their works. Intense!
So I mentioned this lawsuit a few issues back, given the judge involved in the case stated that the suit did not give sufficient grounds for any claim that this work is violating their copyright. The argument being that using the book to train the LLM was transformative in nature, and no proof was given that this has led to an impact on the authors ability to sell their works.
However, the judge refused Anthropic’s request to dismiss the case, given in order to train the LLM, they had a library of books they were using for training on a server. Given they had not purchased the books for that purpose, the judge deemed that piracy! So now they’re facing a fine of up $150,000 in damages per book in said library that they have illegally acquired.
Did I mention their training library had 7 million books in it?
As reported on ArsTechnica, Anthropic is now appealing against the class action itself, given they fail the judge failed to do a “rigorous analysis” of the situation. Instead relying on his experience…???
The petition is here, but the long and short of this 104 page document is this: if they lose the court case, it would bankrupt them, and kill the generative AI industry as it currently exists. Ho-hum…
Excuse me for a moment. The problem with having a tiny violin is when you’re not playing it, it’s prone to being misplaced all too easily…
Another Xbox Studio Head Talks About Generative AI
Ah there we go. I’ve put it somewhere easier to find next time.
So as the headline suggests, another Xbox studio head spoke up on the state of generative AI in the industry. This time it was Brian Fargo, founder of InXile Entertainment, a studio known for the RPG series such as Wasteland and The Bard’s Tale.
In a conversation that covered numerous topics with YouTuber MrMattyPlays, he makes some sound points on using AI to remove the more mundane aspects of craft and let teams focus on their creative processes and efforts. You can hear his thoughts at around 35 minute mark, where he also makes some interesting comments and the value derived from the audience, but also the real ‘threat’ when teams 1/10th the size of his studio are tackling similar sized projects. Good stuff, I enjoyed listening to him.
Wrapping Up
That brings this weeks issue to a close, thank you for reading all the way to the bottom. I hope you’ve seen the value of that which we’re seeking to build, and can help us continue to bring more content to you in the future. Once again, a huge thank you to everyone who supports AI and Games and has done over
In the meantime, I want to give a quick shout-out to everyone who supports AI and Games and has done so over the years. It’s thanks to you we’ve been putting out content now for over 10 years, and want to keep doing what we’re doing at the standard and quality it deserves!
On a total sidebar, a big congratulations to the devs across all of the Battlefield Studios umbrella. I spent a few hours playing the Battlefield 6 open beta at the weekend, and it was pretty fantastic. I also really loved seeing AI bots in the training playlists. It was both great to have them there to remind me how to play Battlefield - I have not really played the series since Battlefield 4 - and to help me feel good and rack up a high score after only a couple of matches.
It was also crushing when I then left those training playlists, and was reminded of how slow my reflexes are when I’m playing against, y’know, actual humans…