22/12/2023 AI to detect fake NFTs and find NFT’s true value: bitsCrunch AMA recap

The NFT market may have taken a downturn, but don't count it out yet. The real question is, are we ready for the next booms and the challenges they bring? bitsCrunch has some solutions.

AI to detect fake NFTs and find NFT’s true value: bitsCrunch AMA recap

Vijay Pravin, founder and CEO of bitsCrunch was the guest of Cointelegraph's latest AMA. bitsCrunch is the blockchain analytics and forensics platform that allows to estimateNFTprices, check NFTs for fraud, get comprehensive NFT analytics and more. The company has already worked with Dapradar, Unstopabble Domains, EY, PricewaterhouseCoopers and Mastercard.

Vijay discussed in detail the bitsCrunch tools available for NFT creators, builders, analytics, brands and enterprises, the state of the NFT market and its challenges.

On NFT market

I think NFTs will make a comeback. If you think back, we had several cycles for the general crypto space. But for NFTs, it was just one pull run. And the market is slowly moving up.

Of course, not all NFTs will make it like we see with cryptocurrencies, but the really good collections will. Look at Pudgy Penguins, which are being sold physically in Walmart stores across the US. Mocaverse, Bored Apes, a lot of blue chip NFTs are also doing really well because they have utility IP associated with them. Established brands also continue to ship into the NFT space, such as Nike, Reebok, Adidas, Mastercard and Visa, Porsche, Reddit, Starbucks, etc.

If you keep looking from a positive perspective, there are still many sought-after NFTs, their floor prices might be around zero at the beginning, but you should look beyond that. Some NFTs are now trading at quite a good volume. Moreover, the moment NFTs are widely used for real-world assets, they will have a strong position in the space.

On wash trading

bitsCrunch is the guardian of the NFT ecosystem: we help prevent malicious activities such aswash trading. The classic example of wash trading is when NFT is bounced back and forth between two wallets. In this case, you never know if it's two different people or two different wallets belonging to the same person. But we have also seen complex patterns that can involve the whole community, ping-ponging an NFT to each other and then back to an original owner.

Different currencies can also be used for wash trades: for example, user A funds user B in Ether, then user B converts it into Lux tokens or other native tokens of some NFT marketplace. It's like credit score: in USA you have credit score, in Germany you have Shufa, in other countries you have some similar credit score.

Some NFT marketplaces and chains have started to implement protection measures, especially the newer ones. They want to be transparent and trustworthy, so they hire forensic teams and look for advanced tools like ours. We have identified more than 10 wash trading patterns on Ethereum, Polygon, Avalanche and Binance. Currently, our accuracy is around 95%, which we are constantly improving.

On NFT analytics

UnleashNFTis our comprehensive NFT analytics and forensics platform. Currently, we support over 30,000 collections across four chains. We select the top ones by floor price, number of trades, and other criteria. The service is completely free and provides complete information on individual NFTs and collections through data visualization and dashboards.

You can see the rarity level, sales history, whether it's a copy or not, and other details. In addition, you can find the market metrics such as market cap, number of trades for a given period of time, number of watch rates that have occurred in a particular collection, etc.

We want to go beyond NFTs because we have a lot of data. We already provide historical prices of over 220,000 collections on Ethereum. So if you want the price of Chainlink back in 2017 or 2018, we are covering that for our premium users. And if you are technical enough, you can already connect your wallet and get access to our API keys with just a few clicks. We've made it as simple as possible so that anyone, anywhere in the world can access the data in a permissionless and easier way.

On AI, NFT prices and copyright violations

To find the most accurate prices, we use AI, which we have had since the beginning. We also look at about 50 key performance indicators, where we look at sales history, watch rate numbers, fake assets, if NFT is a copy from another collection, how often it has been flipped, and so on.

But compared to art, where people can have different perspectives on pricing, NFT's price depends most on the community and the value it provides. Like Crypto Punks and Bored Apes who were early in the game and continue to add value to their communities: you can find restaurants, apparel, even ticketing and a lot of other utilities around these NFTs. And these are not classic brands coming into Web3, these are Web3 native brands that are making huge waves.

I have heard that some major Web2 brands are planning to launch their collections as the market moves up. So I think NFT lending is going to be prominent because these are assets that are likely to be held in a certain balance for a long period of time. In the same way that people have held Bitcoin for years, people will hold NFTs while looking for additional incentives. We are already working with some NFT lending protocols: they need to know fair prices for NFT lending, analytics and forensics to identify "good" and "bad" NFTs.

We also use AI to help brands track and protect theirintellectual property (IP)in the NFT space with our NFT Copyright Protection Service. There was a case with Hermès when they successfully sued the creators of the "Meta Birkin" NFTs for using their brand without permission. This problem can be ongoing because the more NFTs we have, the more complicated it is to track any IP infringement. But with AI, we're able to identify when a brand's logo or other proprietary assets are being used by someone else for commercial purposes and notify both parties.”

Arts

https://cointelegraph.com/news/ai-to-detect-fake-nfts-and-find-nfts-true-value-bitscrunch-ama-recap

Interesting NFTs
Symptom
Symptom --- Vhils Work with Explosives First explored to further emphasize his destructive approach to creation, Vhils' work with explosives should be seen as a natural progression from his original bas-relief carving technique used on walls. According to Vhils, walls retain within their layers something of the history and essence of the city which can be symbolically exposed when he carves through them. By employing explosive charges to reveal these portraits and compositions, Vhils takes this concept to its logical conclusion. This original technique was first employed in 2010 for the Detritos (Detritus) series of art videos as a visual metaphor for the social and economic volatility brought on by the 2008 global financial crisis and especially today in the covid era of 2020. As such, both this and the subsequent series speak of how, at times of turmoil, the veneer of civilization we take for granted can easily explode and bring forth previously tamed expressions of intolerance, extremism, and violence that simmer beneath the surface. As the entire process is not only eruptive but also fast and dramatic, its result can only be truly appreciated on video. Most of these have been shot using a high-speed digital camera which enables the artist to capture the moment at a very low speed that seems to almost freeze the flying debris, creating a slow-motion poetic intensity conducive to reflection.
Fidenza #529
Fidenza is by far my most versatile algorithm to date. Although the program stays focused on structured curves and blocks, the varieties of scale, organization, texture, and color usage it can employ create a wide array of generative possibilities. Additional project feature(s) => Scale:Large, Turbulence:Med, Colors:Luxe-Derived, Have Margin:Yes, Spiral:No, Soft Shapes:Yes, Super Blocks:No, Collision Check:No Overlap, Outlined:No, Shape Angles:Curved, Density:Medium
#3401
By OthersideDeployer
TPunk 3442
By A1F6BF
Everydays: The First 5000 Days
I made a picture from start to finish every single day from May 1st, 2007 - January 7th, 2021. This is every motherfucking one of those pictures.