AI Eye – Cointelegraph Magazine

In the year By 2022, the “meta version” was everywhere, with tens of billions invested in the sector. Largely substance and user-free, the narrative is forgotten by 2023, with speculations aboard the AI ​​train in a dramatic way.

But generative AI is shaping up to be a new way for users to build metaverses themselves — and by metaverses, we mean virtual worlds where creators own a slice of the economy via blockchain and NFTs.

AI allows non-artists to generate images, non-developers to code programs, and basically anyone with a bright idea to create new worlds and games.

That was at least the bright idea of ​​an AI eye for a feature story back in September of last year, but despite interviews with everyone from Sebastien Borget (The Sandbox) to Ilya Polosukin (Nearby) on the subject of building an AI metaverse, the story went nowhere, at the time it was more theory than reality. .

And while it's still too early to call it a confirmed trend, the game meta opposite Alien Worlds has strong plans to make the concept a reality.

In the year A “Dune-meets-cyberpunk” game created by main contributor Dako in 2021 with 50,000 daily users, it features six DAOs known as Planetary Syndicates competing for a share of Trillium (a token, TLM).

The virtual universe, along with a world-building story, was handled by Dune Prequels author Kevin J. Anderson, and according to Dakoko, much of the artwork, from alien races to buildings, vehicles, and landscapes, was created with the help of AI tools. Medium trip.

“We are using AI to generate a lot of legends, pictures, images, texts. This is from Kevin J. We're building on the original Spark we built with Anderson. So we're trying to do that quickly and at scale,” explains DaCoco's Chief Marketing Officer, Kevin Rose.

Andersen's creation of new aliens, new technologies, and new cultures is devoted to a larger lore model called LoreLynx (based on the open source Mistral 7B LLM), which introduces many diverse contributions to the West into a unified canon. GitHub repository using AI and 6,000 token voters. Users can query LLM about the world built so far, analyze the data on the chain, and use it to create storylines.

“We're creating AI tools, proprietary and otherwise, to enable people who can't draw to build art, and people who can't write to expand on this canon, accelerate stories.”

Rose explains that Lynx can't generate art at this stage, but Dacoco is working on different ways to use AI to create a coherent image.

One example of a user contributing to the knowledge and economy of Alien Worlds is Geoff McCabe from comic book and game company Lightning Jobs. He is behind the NFT comic Starblind, which is set in-universe and sold as Polygon NFTs.

For McCabe, the massive user-generated peer coordination of AI and blockchain in the Marvel Cinematic Universe or Star Wars is a huge leap forward.

Think about it, “Star Wars basically tells the whole world that you can come and make stuff for Star Wars now and make money from it – it's just a matter of being consistent,” he says.

“It's premature, so it will be difficult for people to understand the importance of this for three to four years.”

That said, while AI Eye asked for specific examples of AI-generated content, Rose and McCabe provide a handful, including AI-generated characters for Starblind, community-built games using Lynx, and platform-style game creation. It's called the Trillium Mission, which uses NPCs controlled by AIS.

As small as Metavas is in 2022, it still seems more concept than reality. But it's a great concept.

“Essentially we need content centricity, and AI can help make that happen. So we're in the early stages of that,” says Rose.

The easiest way to participate is by going to Alienworld.io or downloading the Battlefleet Armageddon iOS app.

AI is like crypto: DeepMind is CEO.

Speaking of hype, that's one of the big similarities between AI and crypto – the other being that they're both new technologies that most people don't understand.

DeepMind CEO Demis Hassabis worries about these similarities. “[AI] It's going to bring a whole host of incentives, and maybe some exciting and other things that you see in other mixed areas — crypto, or something else,” he told the Financial Times.

Some of them have now gone into AI, which I think is a bit sad. It obscures science and research.

This week's removal of Amazon's Just Walk Out technology from stores is a good example of a token getting ahead of the technology. AI has been introduced to make it easier for customers to pick up items and avoid the checkout, with 1,000 contractors in India hired to evaluate and train the technology to complete 70% of transactions manually.

And while no one doubts that AI is here to stay, its use in professional settings will be limited until the illusions and untrustworthiness issues are ironed out. Some people are now beginning to wonder where all the profits coming from the hundreds of billions invested.

Albert Edwards, chief global strategist at Societe Generale, thinks the AI ​​narrative looks like a bubble pushing stock markets to record highs.

“Every bubble has a compelling narrative,” Societe Generale's chief global strategist wrote in a note to clients.

“The current narrative focuses on waiting for AI-driven corporate profit growth to fully justify current stratospheric prices. We live in the late 1990s. [tech] We've heard bubbles before and roll our eyes to the sky.

AI is saving lives, improving lives

AI Eye reported last year on an MIT team that used AI to find the first new antibiotic in 60 years to treat MRSA, a superbug that's so far resistant to antibiotics.

Now, a team of researchers has published a paper in Nature that formalizes the process. The researchers created a model called “SyntheMol” to easily synthesize new compounds from a chemical space of about 30 billion molecules.

They gave AI a chance to find antibiotic candidates for another antibiotic-resistant bug commonly found in hospitals, Acinetobacter baumannii, aka Iraqibacter. They synthesized and screened 58 of the resulting molecules, yielding “six structurally novel molecules that exhibit antibacterial activity against A. baumannii and a wide variety of other bacterial pathogens.”

In another example of AI saving lives, an AI tool called Mia was tested on its ability to diagnose breast cancer with 10,000 mammograms on real NHS doctors. Not only did the AI ​​run all the tests that showed the signs, it also found 11 very early-stage cancers by just a few millimeters that the real doctors had missed. Breast cancer patients with tumors smaller than 15mm, when detected, have a 90% five-year survival rate.

AI is helping to generate new patients by making beer taste better. Researchers at the KU Leuven University in Belgium spent five years using AI to analyze the chemical properties of 226 of 250 beers and cross-reference the results with professional and amateur beer reviews. They discovered that they could improve any beer by using complex chemical properties that all real-world beer tasters agreed were better.

“Beer really was better,” said study leader Professor Kevin Vestrepen.

Eric Wall vs Bittensor

Blockchain analyst Eric Wall appears to have wiped $90 million from Bittensor's (TAO) market cap by devastating the AI ​​cryptocurrency project on X, once known as the “altcoin killer” and later criticizing the project. They became disasters like Terra Luna and Hex.

Wall used the example of Bittensor's LLM-like “subnet 1” to explain that the project was incredibly wasteful because users sent a request, and 1,000 miners running LLMs sent back the same response in TAO tokens.

Also read: Revised ‘altcoin slayer' Eric Wall on shitposting and Ethereum scaling

“For each request, you have literally a thousand miners, and they're constantly completing the same task,” he said.

“There's no AI magic to make sure a model works correctly. There's nothing stopping miners from copying and correcting each other's responses, chasing their work.”

Bittensor investor Arca posted a blog to combat “misinformation” and that it wasn't actually “just a bunch of chatgpty clones stuck with a worthless memecoin.”

The article suggests that there are different subnets that focus on everything from creating products using decentralized search or cloud services, and each has different incentive mechanisms to encourage contributions.

“Criticizing the entire network based on the perceived weaknesses of Subnet 1 is like evaluating the Ethereum network on a single application.”

The post also took issue with Wall's claims. In fact, only requests for mining purposes are distributed to the entire subnet. User-initiated queries are routed to a single miner or group based on the validator's routing strategy.

However, he admits that “the criticism of the reward system carries some weight.”

After being piled on by TAO maxis, Wall warned that taking him would be like invading Russia.

“I don't care if you live a million. You will all die of starvation and permafrost in these lands, I will destroy every one of you.

All killer no filler AI news

— Fetch, SingularityNET, and Ocean Protocol are joining forces and merging tokens under a new Artificial Superintelligence Alliance. Voting on the ACI merger closes on April 16. Read what Dr. Ben Goertzel calls the journal's “So Deep Deep” on the topic.

– Tether has launched a new AI division and is looking for an engineer and head of research and development. The company's headline says its goal is to “build the next generation of open AI models,” and it's full of phrases like “innovation,” “protecting privacy” and “democracy,” but it doesn't shed much light on why. Stablecoin company is looking for an AI department.

– Degens who don't bother to think of a name or provide an exchange can use AI to unlock a new memecoin token in around 20 seconds.

How to deploy the next billion dollar meme token on @base, world record speed:

-> pic.twitter.com/HEXamC4sp2

— 👨‍🍳.eth (@CroissantEth) April 1, 2024

— Professor Jeremy Nguyen has put together a simple beginner's guide on how to download and run the nomic GPT4All interface and local open source model on a laptop to privately query your own documents. While the better the hardware the better the performance, users have been optimizing open source models for CPUs that can run on low-end hardware like the Raspberry Pi.

– AI can widen the gender pay and job gap. A KPMG survey of 1,048 Australians between the ages of 15 and 24 found that while a third of blokes are very interested and knowledgeable about AI (or so they say), only 24% of Gen Z women are interested and only 20%. They claim to have a good understanding. A separate study at New York University found that large LLMs tended to re-screen job candidates who took maternity leave.

– LLMs can be incredibly persuasive after being given personal information about them. A pre-published study found that AIs are 81% more likely to convince people of an opinion than human opponents. The results suggest that personalized targeting of individuals with political and commercial ads will be a concern in the future.

Subscribe

A very engaging read in Blockchain. It is given once a week.

Andrew Fenton

Based in Melbourne, Andrew Fenton is a journalist and editor covering cryptocurrency and blockchain. He has worked as a film journalist for News Corp Australia, SA Wind and national entertainment writer for Melbourne Weekly.

Follow the author @andrewfenton