From a painting straight out of the Renaissance generated by a selfie to fully written stories with only a few words, AI is taking the world by storm.
AI seems to be the new buzzword everyone is throwing around with the recent rise in popularity thanks to ChatGPT, DALL-E, and Midjourney, just to name a few. AI has officially gone mainstream and is becoming widely used among Gen Z and Millennials. New AI tools seem to be popping up everywhere. It can be overwhelming to sort through them all and know which ones are legitimate and valuable to use.
There seems to be quite a few questions around why we still need web3 when we have AI.
AI is an incredibly powerful tool, first and foremost, but many issues still remain with web2 including: large amounts of centralization and private code, at times lack of transparency, and a number of barriers to entry for financial freedom.
Four reasons why even during crypto winter, I remain bullish on blockchain and web3:
Blockchain technology is fast, transparent, and verifiable.
It allows for a constant cycle of innovation by not reinventing the wheel and pushes for open source development
There’s more financial freedom to easily send and receive money in seconds without all the hassles of traditional banking restrictions
More individual ownership over data
“Over 61,000 new developers wrote open-source crypto code in 2022,” according to Electric Capital.
Despite the ongoing bear market coupled with the chaos of FTX, more developers than ever before are building in web3.
These are just a few of the reasons why we still need web3. The innovation in both blockchain and AI is groundbreaking and disrupting industries across the board.
However, the combination of the two could be the catalyst for removing barriers to entry for both AI and blockchain. This combo can dispose of antiquated ideas and systems to make way for a new era of transparency, efficiency, and individual financial freedom and data ownership.
So, what is AI exactly?
Oxford Dictionary defines Artificial Intelligence (AI) as, “the theory and development of computer systems able to perform tasks that normally require human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages.”
There are different types of AI including Machine Learning and Deep Learning. Levity illustrates and explains the relationship between the three below in more depth.
As seen in the graphic from G2, Machine Learning includes virtual assistants like Amazon’s Alexa. We have now started to reach the Machine Intelligence stage. Unlike the characters in Westworld on HBO, machines have not gained consciousness…yet.
New and disruptive technology like AI is always going to cause a lot of chatter, controversy, and interest among the masses. AI has existed for decades, but has gotten more advanced. Now because of applications like ChatGPT, the everyday user can now utilize and play around with AI technology that traditionally wasn’t previously available to the public.
Whether we realized it or not, AI has been an integral part of our everyday lives even before the recent boom. From Amazon’s Alexa to the Roomba, AI has been a constant in daily life without the majority of people knowing it was there. Even Uber uses AI to help with detecting fraud and determining the best pick up and drop off locations. UberEats uses it to generate targeted restaurant recommendations and accurately estimate wait times for restaurants.
The biggest mistake a company can make right now is dismissing the technology completely. The companies that didn’t survive the 2000s were the ones that didn’t adapt to the times and embrace new ideas and technology, like Kodak and Blockbuster.
Should AI replace your entire marketing team or all of your writers? In my opinion, no. However, using it to make work more efficient, can increase the amount of work you can get done in a day and also improve the quality of the work.
Why work harder, when you can work smarter with AI?
AI Market Size
Precedence Research estimated the size of the AI market to be around $119 billion in 2022 and is predicted to grow to over $1.5 trillion by 2030.
Some highlights from recent raises in AI:
ChatGPT gets a $10 billion investment from Microsoft. ChatGPT is part of the OpenAI family and uses prompts to create generative text. Jasper raised $125 million in a Series A with a $1.5 billion valuation. Similar to ChatGPT & DALL•E, Jasper uses Generative AI to produce AI images and writing. Stability AI is also producing generative AI images and raised $101 million with a $1 billion valuation.
Raises for AI in Web3
We are also seeing large funding rounds for web3 companies integrating AI even with the bear market. Plai Labs raised a $32 million seed round led by a16z in January 2023. They are building decentralized social platforms combining AI and web3.
One trend in Venture Capital I don’t see ending anytime soon, is VCs investing in web3 infrastructure. Throughout the bull and bear market, we have continued to see a steady influx of new infrastructure companies in their portfolios. For example, Nillion closed a raise of $20 million at the end of 2022. Nillion is a web3 infrastructure company building a decentralized network that isn’t using blockchain technology, but still incorporates the decentralization aspect of web3. In addition, Spice.ai raised a $13.5 million round to provide tools to developers to build blockchain apps using AI.
AI Venture Funds
We are seeing more venture funds focused on investing in AI. According to BetaKit, Radicle Ventures launched a new $550 million AI Fund. In addition, Modus, a NY based venture platform, expands to Africa with their $75 million AI & Blockchain Fund. Back in September 2022, we saw Digital Transformation Capital Partners (DTCP) launch a $300 million fund for AI, web3, cybersecurity and fintech startups that they are to finish fundraising for sometime in 2023.
I predict we will see less funding than we saw in 2021 and 2022 in general, but are still focused on finding needles in a haystack. AI has been heavily funded the last few years and has garnered immense buzz across the globe, so I don’t think we will see AI funding drop significantly. Although, with all the buzz comes saturation, so finding those key teams and innovative use cases for AI is critical.
Generative AI Portfolios
Dealroom’s infographic provides a glimpse into the Generative AI portfolios of some of the top VCs like Sequoia, Lightspeed, a16z, Microsoft, Tiger Global, etc.
Issues with AI
Some issues that are currently facing AI include copyright, bias, centralization, and the loss of jobs due to automation.
Copyright
According to an article in TechCrunch, there’s a massive gray area when it comes to the IP rights for images and/or text created from scraping data from the internet that is not all open source and has copyright. There’s also ambiguity around a computer generating images or text, because who would be held liable if they infringe on the copyright?
In January 2023, Getty Images sued Stable AI over alleged copyright infringement.
The solution here might not be as clear cut as I’m alluding to, but by having a platform that only aggregates open source and open license content instead of scraping from all sources is a better route to go to combat copyright infringement.
Bias & Discrimination
With AI, comes bias and discrimination. Stanford reported in their AI Index Report 2022, “The systems [AI] are growing significantly more capable over time, though as they increase in capabilities, so does the potential severity of their biases.”
Forbes finds that with human bias also comes discrimination and bias based on gender, race, and age in AI algorithms. There’s also a lack of diversity in software engineers that specialize in AI as we can see in the chart below. These are the people teaching AI, so this is the first critical step in my opinion to having fair and ethical AI.
Just like with any education system, we need a diverse set of teachers coupled with a curriculum that is inclusive and unbiased.
Centralization
A lot of AI is largely centralized and proprietary (built on closed source development) by some of the largest tech companies in the world like IBM, Microsoft, Amazon, Apple, Google, etc. These are the same companies with vast amounts of data and use their data sets to teach the AI. This can be problematic because they have their own interests in mind and those interests aren’t always in line with the needs of the overarching community.
Loss of Jobs due to Automation
Ironically, there used to be people solving complex math equations whose job title was, in fact, "Computer". The technology went on to replace the position, but new jobs and positions came with it and opened up a whole new world of possibilities. So, while AI will potentially eliminate a lot of positions held by people, it will also open up new job possibilities and positions. AI will stimulate the world economy and reshape the world around us much like the computer and the smartphone did.
Latest News in AI
AI is disrupting industries across the board and is changing how we do customer service, marketing, sales, coding, and art, but we will see it in the future be used to disrupt other areas. Here are some highlights of AI shaking things up:
As we can see in the LinkedIn post from Linas above, ChatGPT achieved one million users in one week. What took ChatGPT only seven days, took the other tech giants months, even years to accomplish.
Only one month into 2023 and the founders of Instragram announced their new company, Artifact. Artifact is a news feed app powered by AI that operates in a similar way as Tiktok.
As pictured below from the Global Head of Luxury chez Snap Inc, Geoffrey Perez, shares on his LinkedIn about Louis Vuitton entering the metaverse with a filter that uses AI on Snapchat.
Not only is AI a game changer, but it’s also an incredibly intelligent student that has passed the bar exam, an MBA exam, and a medical exam. Will I be the first patient to get operated on by an AI powered robot doctor? No thank you, but there could be a significant number of medical advances that can happen with AI in the healthcare system.
The graphic below shows the 2023 Web3 + AI Market Landscape. This visual highlights just how many Web3 companies and startups are integrating AI.
Value of Integrating AI & Blockchain
The opportunity for AI lies in the ability to not only disrupt web2, but enormously improve some of the key issues in web3. An article from Forbes Digital Assets even goes as far as to say, “But AI and crypto, together? There’s an untapped potential we’re only just beginning to see put to use. It could transform both industries.”
Interoperability
One of the key issues with crypto and web3 is that we need to have better interoperability. Interoperability makes it easier to move assets from chain to chain and allows ecosystems and tech stacks to interact with each other in a more composable way than web2 and traditional finance. Web3 has not reached this level of seamless interoperability where all complexities of different chains and their ecosystems have been stripped away for the end user.
In an article from Coindesk, Haseeb Qureshi (Managing Partner at Dragonfly Capital) talks about the five biggest problems facing web3 with one of them being, interoperability.
“Right now, you're extremely aware of what chain you're on. You know if you’re on Solana or Avalanche or Ethereum. At some point in the future, almost without a doubt, you will have digital assets and you will interact with applications – and that will be your relationship with [the blockchain].”
So, how can AI solve the interoperability problem? Qureshi goes on to say, “In the future, if you want to do yield aggregation, instead of just looking at Yearn, you’ll look at what's the best yield anywhere in crypto.”
Imagine if you had an AI tool that could, with a few prompts, find the highest yield across all blockchains that can filter for specific risk profiles according to the user’s preferences.
Improved Security
As many crypto users may have been scammed or had their wallet drained at some point in time, improved security is critical for global adoption. By integrating AI into dApps and DeFi, AI can make them operate more efficiently and securely.
As we can see from the research from Token Terminal below, 50% of all exploits in DeFi are from cross-chain bridges with around $2.5 billion in stolen funds since September 2020.
AI could not only optimize for the best yield, but also could scan through open source code to detect potential risks in DeFi.
Potential bugs could be found and fixed based upon past instances of cross-chain bridging exploits and other generally safe coding practices. This could be critical since a lot of smart contracts utilize open source code from other protocols and dApps increasing dependent risks.
According to Chain, a web3 infrastructure provider, “Through identifying and detecting malicious activities on the network, AI can detect patterns of suspicious transactions. AI can act as an added layer of security on the blockchain, providing automated solutions for security, scalability and privacy concerns when utilizing a blockchain network.”
Certik, one of the top companies for web3 security audits, is already using AI for one of their products, Skynet. According to their website, Skynet powers on-chain monitoring and performs continuous AI-based scanning of your contract code for vulnerabilities using Certik’s growing database.
Another related use case I see for AI in web3, is a tool to detect flaws in development of smart contracts in real time as it’s being written. This could put a stop to hacks and exploits before they even go live in the first place.
With the amount of exploits we have seen in DeFi and in bridging assets, integrating AI into web3 could be instrumental in preventing future hacks. This would also help bring more people into crypto who have been skeptical of the security and safety of their funds and tokens.
Upgrading the Metaverse and Gaming
We have yet to see a game or metaverse in web2 or web3 that even closely resembles the world in Ready Player One. AI can be used to improve gaming graphics and create more addictive games that people want to play.
Traditional gaming is heavily centralized and most of their code, along with the AI, are generally not open source. However, gaming is a huge market and could be the way to onboard one billion users if a web3 game takes off. AI could help draw in more users to the user base if implemented successfully.
Sarah, a Research Analyst from Ancient8, gave more insight on how AI can significantly decrease development time for NPCs and storylines: “AI allows NPCs to get smarter and respond to in-game conditions in novel and unique ways as the game progresses. Utilizing this technique will profoundly reduce the development time of NPCs, as hard coding their behavior is a tedious and lengthy process. This is especially important for web3 games that are building as fast as they can to keep their communities interested and engaged.”
Here are some companies currently combining web3 and AI for gaming and the metaverse:
Today, a simulation game using AI & web3, was built by renowned gaming studio, Mr. Kite. The team is comprised of past employees of Unity, Nintendo, Epic, Blizzard, Sega and even previous Oscar winners.
Kinetix, featured in the chart above from a16z, is a platform to create and mint animations into custom NFTs or “Emotes” for avatars in web3 to express life like movements and personalities from existing videos.
OpenFabric is leading the way for AI in the metaverse and gaming. They are described as, “a platform for democratizing access to AI that allows anyone to create, share, and monetize AI services at scale.” Currently, AI takes an immense amount of CPUs and is expensive to create and/or access. OpenFabric makes it simple and affordable with their product, HyperAI. By using HyperAI, it creates AI generated videos from a text prompt.
The Bornless is a first person shooter (or FPS) horror game that is completely free to play. It is supported by AWS Activate and Nvidia Inception Program. They built a prototype for generating in-game assets with AI.
Render Network by OTOY is revolutionizing design and development tools for artists and creators for video games and the metaverse using web3 and AI, something that OTOY calls AI neural rendering.
Other Interesting Integrations
Digital Art Authentication
Numbers Protocol is a web3 network for decentralized photos that is solving the issue of centralization, security, and authenticity for AI generated images and digital collectibles. They also were selected as a finalist for SXSW in Austin this year.
Sleep-to-Earn
Dormint is using web3 and “AI powered sleep analysis” to help people sleep better and earn tokens through gamification. Now, this sounds like something I could be very good at. All jokes aside, it’s an interesting way to see AI integrated because of the potential possibilities going forward for improving sleeping patterns and maybe one day helping combat insomnia.
Research
Buildspace posted about ChatGPT and how it can read CSV files from Excel, so a lot of customer research can now be done for you without having to comb through spreadsheets or long research reports.
This could easily be leveraged into web3 to offer better tools to your B2B customers based on their own data and usage.
More Use Cases to Come
Recommendations Tool
We are already starting to see AI used to make recommendations to consumers on what to buy based on buying and search histories. What if someone built a Recommendation tool for web3?
The tool could be used to make recommendations for UI/UX upgrades based on top-rated apps UI that their customers are using based on wallet activity on chain and sentiment on ease of use on Twitter and other social media platform
Ads Tool Powered by AI and Web3
Design and implement a tool to create AI targeted ads — AI could be used to place more targeted advertising in front of the right users by leveraging web3 tech. You could have better targeting of ads based on each user’s cookies, search history, combined with their interactions of NFTs, trades, or other activities on-chain. This would be an interesting way to bring more traditional web2 brands into the space to reach new and existing customers.
Summary
Overall, blockchain and AI are both non-competitive, complementary tech stacks that can help enhance and create new products together to enhance our world. Combined, they can balance each other and be more applicable for everyday use cases. Web3 helps AI by being more decentralized with its ethos of open source development and strips away some of the hassles that come with traditional finance. AI can make web3 more efficient, improve graphics for gaming and the metaverse, and bring new users to web3.
About the Author
Kelli graduated from one of the top business schools in the USA and was voted 21 under 21 from the Tippie School of Business. She has 10+ years of combined business development experience from both the traditional finance world and web3. She most recently worked at The Giving Block in crypto partnerships. Kelli is passionate about emerging technologies such as web3, digital collectibles, and AI, and is looking to break into venture capital. In her free time, she enjoys traveling, reading, learning about new developments in tech, and making mosaics.
Very well written with lots of insight on AI & blockchain, well done