Hello everyone! Now that we’ve all deeply marinated in 2022’s juices and allowed it to char our flesh, we are just about ready to deal with the butt-clenching horrors it has to throw in our general direction. Never mind though! Because something totally new is coming: it’s called ‘next year’. Forget everything you learned about 2022, and start cramming for 2023 because honestly it’s probably going to destroy us all (lol, oh well).

Today as a special end of year treat I decided to engage the knowledge of four industry experts, and ask them to reflect on the biggest events on 2022, and share their predictions for 2023. Those experts are, in no particular order:

• Rosa Carbo-Mascarell: a game designer currently working for Improbable. Today Rosa tells us what a metaverse really is.

• Clement Briens: works in cyber intelligence, and provides us with a concise history of how politically motivated cyber activity has changed (or not) over the years.

• Lucie Kaffee: is an NLP researcher, who shares some hard truths about ChatGPT

• Ben Whitelaw: is an expert in content moderation, and will talk briefly about the role CDN’s play in moderating content online.

Keep reading to learn more about who these people are and what they think about the past AND the future. Please note that responses have been edited for clarity and length.

⏮️ What happened?! Looking back on 2022

The year in technology has been like a nauseating roller coaster that has been impossible to get off, no matter how many keywords you mute. As such, I asked the experts what Big Thing stood out most of them.

Lucie described this year as “chaos” with the most entertaining part of the chaos being Elon Musk’s Twitter deal. She’s obviously right — literally no one can argue with this. In fact Rosa said that her favourite moment of the entire year was “when Elon Musk launched Twitter Blue and a bunch of people impersonated businesses. Sean Morrow in particular is a hero for that Eli Lilly tweet.”

Twitter’s ever widening fountain of discord this year may have actually given regulators the motivation they need to get up and do something. Ben mentioned that his biggest surprise in 2022 was “the assertiveness of regulators, and the way that they have flexed their muscles,” explaining that Thierry Breton from the European Commission was “quick to tell Elon Musk that ‘the bird will fly by our rules’ and has not minced his words on several occasions since. He’s confident that the Digital Services Act will fix the multitude of platform problems that we have come to know and isn't afraid who knows it.” I am also very interested to see what effects the Digital Services Act will have on all this — I have written about what its rules are here.

The chaos that Lucie was referring to obviously spans beyond Twitter, and also spans way beyond being just a bit of entertainment: “In previous years, it seemed like there was steady progress and an overall understanding that tech should grow and evolve in reasonable ways — and ultimately we could tell ourselves that we are still trying to better humanity.” In 2022, however, Lucie saw this fever dream of hope kind of melt away as huge swathes of staff were let go from Big Tech companies. For Lucie, the huge layoffs marked the end of “the stable upwards way of Big Tech companies” and are “an indicator things might change on a large scale”.

Reading the year from a different perspective, Clément referred to 2022 as a “golden age of open-source intelligence, with armchair analysts from all over the world suddenly becoming assets to Ukraine by debunking Russian disinformation, tracking troop movements, and documenting war crimes.” Ah, war… the ultimate manifestation of chaos. Something that stood out to Clément this year was the use of new tech to aid war, such as “Deepfakes actually being used effectively to social engineer US officials (e.g. impersonating Michael McFaul, former US ambassador to Russia) and disinform Ukrainian civilians (e.g. Zelensky deepfakes on hacked UA media websites).”

When we reflect on the year gone by, I often find it’s nice to end on a positive note — and I have confidence that by this time next year, I will have found one.

🔍 Let’s get more specific — and start looking forward (which is the direction that time goes in…)

I like to treat my experts with dignity, so I also sent over questions that were more specific to their respective fields. In every answer, there is at least one thing each for you to memorise and pass off as your own opinion in your next conversation (I’m not saying you should do that but you know, you COULD).

Clément Briens: “Hacktivism will follow the ebb and flow of conflicts in the meatspace”

Clément works in cyber threat intelligence, researches disinformation in their spare time, and is obsessed with open-source intelligence. They're also running MisinfosecUK, a community of disinfo researchers.

In January 2022, before Russia invaded Ukraine, a hacker group named Belarusian Cyber Partisans did a ransomware attack on the Belarusian train system. For the first time in the history of ransomware, they were not demanding money — they were demanding the release of political prisoners. There are also engineers out there who have been offering their cyber intelligence services for free to the Ukrainian government in order to help the war effort. Do you expect that we might see more politically motivated (whatever that means) ‘cyber crime’ following all this?

Hacktivism has been around for a while, and has gone a little quiet over the last couple of years. We’re only recently seeing a resurgence now because of the Russian invasion of Ukraine, as well as political unrest in other countries like Iran. The golden age of hacktivism was arguably during the Occupy Wall Street movement, when negative sentiment against financial institutions and the ‘establishment’ was high.

Hacktivism will likely follow the ebb and flow of social unrest and conflicts in meatspace, so hacktivist groups on both sides of the Russia/Ukraine conflict will probably keep at it as the conflict continues. However one thing to note is that the tradecraft of hacktivist groups hasn't evolved significantly in the last decade or so; they either focus on disabling their targets' public infrastructure (via defacements or DDoS) or using ‘hack and leak’ tactics to damage their opponents.

My theory on why this hasn't changed in the last decade is this: hackers with sufficient skills to ransom organisations or exfiltrate sensitive data are more likely to do so for profit, given the success of Ransomware-as-a-Service (RaaS) gangs like Conti or Lockbit. Kids with sufficient skills that would have joined hacktivist groups like Anonymous or Lulzsec back in 2010 are nowadays more likely to emulate groups like LAPSUS$ (extorting Microsoft, Nvidia, Uber, allegedly led by a 17-year old) in pursuit of profit rather than patriotic aims.

🔮 Clément’s predictions for 2023:

• $TSLA will zero out and collapse, throwing the entire electric vehicle industry back decades

• The security industry bubble will pop, meaning a lot of security companies will close shop; their customers will be using outdated products, leading to more data breaches

• Some engineer will create a rogue AI (think chatGPT connected to the internet) that will break the internet for a couple of days

• EA will collapse as an organisation due to the end of Future Fund, proof that AI safety is hot air, leading to a balkanisation into new movements

Lucie Kaffee: “how much more convenient could you possibly make life?”

Lucie is a postdoctoral research fellow at the university of Copenhagen, working on the intersection of multilingual NLP, knowledge graphs, and whatever happens to be the most interesting to her at the moment (right now: cross-cultural values). As a Wikipedian at heart, she tries to tie in her work with the online encyclopedia.

Well… this year internet users have, whether they’ve wanted to or not, been baring witness to a galaxy-scale explosion of generative AI tools. From DALL-E to ChatGPT to Stable Diffusion to whatever else. My questions are simple to ask, but potentially complex to answer. Why is this happening now? E.g. what ‘problem’ is something like ChatGPT really solving?

Against my own convictions, I have to admit I was quite impressed with the quality of ChatGPT's answers to my silly prompts. Is there a possibility for a model like this to help people who are not native English speakers express themselves better? Sure (imagine how much more eloquent this text could have been formulated if it was post-edited by ChatGPT).

But the reality is, as long as there is limited access, and the details of the system are only known to team members who work under NDA’s, we will never find out what could be possible. So for now it is just a very big, very expensive, and mildly impressive toy, that makes nerds go: ah, wow (all lowercase).

Systems like ChatGPT take a lot of money and resources to build. From your perspective, is there something else in the field of AI on which this money and resource may be better spent?

I keep hoping for outreach to communities that are currently not represented in the tech world so that they get to have their say in what we should spend money on. All the Elon Musks in the world can burn money all they want, but the impact is minimal if they spend it on themselves. How much more convenient could you possibly make life? It's going to get really fun when we start working on problems that are hard to solve because we have to change our approach completely. For example, because of a lack of training data. A lot of the impressive work is technically incredible but has yet to be proven to show how it can support us all, as a society.

🔮 Lucie’s predictions for 2023: We will get more closed-source AI technology, which no one really knows what to do with — which is a shame! Why not let everyone play with all the models?More chaos in the ‘traditional’ and big tech companies, which could leave a new void. But maybe that void gives the possibility for some smaller, really cool companies to take off. It seems I still have some hope!

Rosa Carbo-Mascarell: “ Everything already is a metaverse. It's metaverses all the way down”

Rosa Carbo-Mascarell is Design Manager at Improbable. Designing for multiplayer and social is her specialty. Before, she worked as Lead Game Designer and LoveShark, Nexus Studios, Milo Creative, and others, mainly making games for change.

It seems that the word ‘metaverse’ has very quietly phased out of our viewports before it ever had the chance to turn into anything useful — but maybe there are things bubbling under the surface I am just not aware of. From the perspective of someone who’s been working in games for years, and now works at Improbable, who have arguably been making ‘metaverse’ products since before the buzz word hit the mainstream, why has so much money suddenly been poured into this, with seemingly little point? E.g. the EU recently hosted a 387k ‘metaverse party’ and only five people showed up.

Sometimes rich white dudes decide what they want to see in the world and start bankrolling. Then, the bubble they're in bursts and it's onto the next buzzword. This happens in an endlessly repeating cycle. The metaverse is just the new word being rolled around to katamari in investors. The funny thing is, metaverses have always existed in one form or another. They're called games.

What kinds of products/experiences already exist that you feel can be likened to what people mean when they say ‘metaverse’ these days? This could be something you’ve seen at Improbable, or not.

When LoveShark went bust and I was looking for a new job as a game designer for multiplayer games, I quickly realised that ‘metaverse’ was just the word that a lot of places were now using to mean ‘multiplayer games’. Everything was a ‘insert brand name here’-verse.

To go a bit academicy for a sec, in the early 20th century, Johan Huizinga was already talking about metaverses. He called it “the magic circle”. The magic circle is a world we all enter when we play a game where there are new rules and new ways of seeing the world. When you're a kid and play the floor is lava you're entering a magic circle where the floor isn't just made of wood or carpet, it's lava and if you touch it you die. By that definition games are already metaverses and they have been since we first started playing The Game of Ur.

If you want to go further, Huizinga then goes on to say civilisation was built because of play. Law, religion, money, rituals, are all magic circles/metaverses that we all agree to play by. Everything already is a metaverse. It's metaverses all the way down.

🔮 Rosa’s predictions for 2023: A lot of companies are going to go bust in 2023 since investor money has been drying up. A lot of game companies have gone into a hiring freeze as they try to weather the storm. I think late by 2023 things will be looking up. In the meantime hold on, layer up, donate to your local food bank, and support the strikes.

Also, hopefully, we'll move on from ‘metaverse’ as a buzzword.

Ben Whitelaw: “once there’s evidence of real-world violence, Cloudflare’s argument crumbles like an English penalty taker at a major football tournament”

Ben currently works for FT Strategies, the Financial Times' subscription consultancy, helping media organisations grow and become sustainable. Previously, he worked for the European Journalism Centre, The Times and The Guardian. He writes Everything in Moderation in his spare time.

I think this year has seen a lot of chat about content moderation in many directions — and maybe that chat hasn’t been the most nuanced. But what stood out for me was Kiwi Farms eventually being taken down by Cloudflare. I’d say that before this incident, I was never 100% clear on where I stood when it comes to CDN providers and other ‘web infrastructure’ services picking and choosing who gets to have a website. But, Kiwi Farms is a site that actively coordinates targeted harm to individuals — so it really shouldn’t exist. Cloudflare initially didn’t take Kiwi Farms down because they thought that would count as them taking a political stance — do you think CDN providers et al should actively avoid ‘being political’ as so many of these companies put it?

Well, the first thing to say is that Cloudflare has a history of doing elaborate dances around decisions related to the websites that it provides services for (see 8chan and The Daily Stormer). Each time, it writes a long blog post arguing that it is merely infrastructure and shouldn't play a role in deciding what people can or can't say. Then, there's evidence that such speech led to real-world violence, and their argument crumbles like an English penalty taker at a major football tournament. Kiwi Farms was a mini-drama in the same vein.

Would you even consider what Cloudflare did an act of content moderation?

Infrastructure providers like Cloudflare clearly don't want to be seen playing a kind of content moderator role, but they retain the right to do so in their terms and conditions. More often than not, they have dedicated trust and safety teams of their own to set policies and make calls, just like platforms. In that sense, they are what Ben Thompson calls "the backstop to user-facing platforms that refuse to do their job". I think that neatly sums up where we are right now with CDNs; they are an important layer in the internet's moderation stack but still only one layer. Whether or not exercising that power that counts as ‘political’ is for others to decide.

🔮 Ben’s predictions for 2023: News publishers will continue to fail the general public by covering online speech issues solely through a technology lens, and in broad strokes. I’d like to see more media outlets recognise that technology policy is as directly relevant to their readers and viewers as, for example, health or crime, both of which are far better staffed. We need deeper reporting and smarter analysis than simply ‘mad billionaire liked a right-wing commentators tweet’. It’s why efforts like Tech Policy Press, New_Public, The Information Ecologist, Reboot and of course Horrific/Terrific are important.

🥨 That was delicious, thank you

I just want to thank everyone for their responses, and also thank my readers for even caring enough to open these emails. I hope today’s was especially interesting and useful to you.

If you enjoyed this issue and can afford it, think about maybe updating your subscription to a paid one — Horrific/Terrific is something I do in my spare time, so it would be great to get some money for it. I won’t lie, you won’t get that much extra stuff for paying, because it’s already a lot of work writing a weekly newsletter all on my own. But look at it this way: if more of you start paying me, I will be able to make more content. THANK YOU.

✨If you sign up using this very link, you will get a 20% discount. I’m so nice! The link is valid up to the 6th of Feb so hurry up.✨