Updated: Mar 30
This week we're focusing on Unreal Engine, Ultima Online, CBDCs, ChatGPT 4, runaway AI, light-based optical computing, and more!
“Welcome to the 49th edition of Spatial News™! Please take off your shoes before you enter, but leave your socks on. Nobody wants to see your feet.” — Joh
Game developers can use
to generate environments just like that <snapping fingers>. Theo Priestley takes it even further.
“[B]ut what if you could use player behavioural analytics and artificial intelligence to create unique experiences for every player? [A] single game could evolve over time infinitely without the need for downloadable content to be developed by the studio. Generative AI and GPT4 models could create new storylines and quests, new narratives, new NPC dialogues.”
He later explained in the comments that “Players won't need to input text, their actions and play style will inform.”
My first idea for a game company was pretty much this: games, informed by Perceptual Control Theory’s negative feedback loops, that adapt to a player's gameplay (see SN "#021). Naturally, I love the idea.
In related Unreal news, it also seems like Tim Sweeney’s “master of the metaverse” plan is coming together.
Virtual world hopping with digital assets in tow (Image courtesy of Joh Orengo/DALL-E 2 and the scraped art of unknown artists)
Another crazy read.
"You could kill someone's pet, skin it and give it to them [as meat]. Why would you do that?"
True, but why would you design that into into the game in the first place?
Gross in, gross out.
Just remember that there are “all sorts of possibilities for peaceful protest”, including
“a more global adoption of cryptocurrencies when people realize the downsides of money with track-and-trace, selective purchase censorship, and expiry dates[…]”
Not to mention all of the hacking.
Last newsletter I voiced my concern that cryptocurrencies lead to CBDCs, but Keir brought the Pepsi Twist to the party.
I want to believe that democracy isn’t doomed like the Orthoverse…
You already knew that GPT-4 is multimodal now (you can enter different kinds of input like video, sound, and images besides just text). Upgraded as it is, it’s still “fundamentally flawed”. It can “still generate biased, false, and hateful text; it can also still be hacked to bypass its guardrails.”
“OpenAI has allowed its stunning ChatGPT AI to reach out into the world with staggering new powers. It can now access the internet, run its own code to solve problems, accept and work on uploaded files, and write its own interfaces to third-party apps [like Zapier which itself can connect to some 5,000 apps].”
♫ ♪ Learn, baby, learn! This go inferno! ♫ ♪
Back in SN #043 (Dec. 15th, 2022 to be exact), concerning AI ethics, I wrote,
“Here’s one way to be ethical with AI, DON’T CONNECT ONE TO THE INTERNET or else we’ll probably lose control pretty quickly”.
Great, now, I’ll have to push the goalpost further back: DON’T CONNECT ANOTHER AI TO THE INTERNET, TOO. The last thing we need is two AIs going all “Neuromancer”.
Image courtesy of the Brazilian publisher Aleph (illustrated by Death Burger)
This just in! An
with signatures from 900 of tech’s titans and notables, including Elon, Woz, and ‘em, has urged a 6-month moratorium on training LLMs so that guard rails can be built before things get out of hand.
Government leaders need to be informed and regulations need to be put in place. True. We don't understand how black box AI works and where all of this will lead. At the same time, we might already be too late (where were the futurists leading up to GPT-4?).
Theo Cosmora comments (in an al XandR Bagg LI post where I first saw Time’s article) “Money, control and extreme competition will unfortunately determine this outcome, not [conscience] and ethics, which are dumb naive gobbledegook to some in elite tech as we know.
Plus you'll have some parties asking why should they stop when the Chinese and Russian governments are clearly not going to be stopping, which is not a ridiculous question when you think about it. What's the answer to this last question?”
Another reason is system mechanics. A developer buddy, who shall remain nameless, explained it to me like this, “It takes too long to implement a change in complex systems due to the slight delays in feedback loops. By the time the change hits, the system has already bypassed the change. AI is running away at a pace that they won’t be able to catch.”
The jinn is out of the bottle.
Over/under 5 years things go “end of days”. Who wants to take that bet?
Forget it, by then ‘your’ digital money will probably get frozen by the powers that be for you even suggesting that it might actually be the 'end times'.
Here’s some light news.
“Imagine a home computer operating 1 million times faster than the most expensive hardware on the market. Now imagine that level of computing power as the industry standard. University of Arizona researchers hope to pave the way for that reality using light-based optical computing, a marked improvement from the semiconductor-based transistors that currently run the world.”
Hot diggity dog!
(Thanks for sharing, al XandR Bagg.)
Oh no… but what’ll happen when you place your light-based smartphone into your pocket black hole?
(Ha! That GIF never gets old.)
<Your Spatial AD Here>
Immersal is seeking partners to collaborate with us in creating immersive content such as AR art, AR games, and other location-based experiences using our cutting-edge Visual Positioning System (VPS) technology. Ready-made Immersal City-scale VPS maps will be released for Helsinki in March 2023, followed by Tokyo, Seoul, and Los Angeles in 1H 2023.
Follow the link to get access to ready-made city-scale map!
Sponsored by Immersal
Introducing Shin Software
When in 2011, under the wise guidance of Stefano Provenzano, we decided to create Shin Software, we had clear in mind our mission: bring the interactive, expressive, and engaging power of video games graphics into the business world.
A lot has changed in more than a decade. Both the market and companies have become more aware of the benefits and opportunities of real-time interactive 3D apps and the computational power of devices has faithfully followed what Moore's first law theorized, opening the door to previously unthinkable scenarios and allowing us to release data-driven solutions for specific vertical verticals (fashion, machinery and manufacturing) capable of managing a huge amount of variables in real time.
Somebody said that "Hard work pays well" and that's why our R&D department has continued to develop new features to fully exploit the potential of emerging technologies.
In this context and strengthened by the robustness of our SHOWin3D platform, it was possible to give birth to the "Zegna X" project, which with its 48 billion product variations is going to represent a game changer for the UX of the fashion industry.
SHOWin3D's fashion configurator
We are confident that the communication strength of a network like Spatial8 can stimulate other companies and accelerate the digital transformation process for the benefit of the entire industry.
Sponsored by Shin Software
Thanks for getting caught up in the frenzy with us, Spatialists!