In what looks like a major shift in investment focus by VCs, artificial intelligence is now a leading narrative and focus for most VC funding.
Currently, I'd admit, as most people may already acknowledge, the concept of AI on blockchains just sounds like a million bots running up chain resources and potentially leaving around lots and lots of trash in the process.
Who in their right sense would be appreciative of a technology that creats wastes consistently just to function?
Yes, it's absurd when you think about it, but that's only because we largely focus on short-term realities, ignoring what the implications may be long term.
To clarify, AIs have to create wastes and make mistakes to learn and grow, it's literally same as how humans improve, just faster.
If we want to look at the numbers, there's probably over 600 blockchain projects with AI related solutions that's been funded by VC capital in the past 6 months as AI leads the narratives by 30.06% of VC investment focus.
At least that's what data from Cryptorank currently suggests.
In terms of value, well, there's explicitly not a comprehensive data on just how much has gone to AI-blockchain projects, but considering that just looking through funding rounds, there's the common scenario of AI-blockchain projects receiving up to $10 million for each round.
With our estimates of over 600 projects being funded in the past 6 months, we are effectively looking at $600M to over $1 billion going to AI x Blockchain projects.
Are AI-blockchain Projects even that special?
Many are making the mistake of downplaying what AI will be capable of doing in the next couple of decades.
I guess this is a common reaction people have with new technology, it happened with Bitcoin(crypto) and now, AI is facing the same reaction.
You hear things like “It will never be smart enough” or “It will never perfect image and video generation”
Is that true though? Simply looking at the evolution of image generation solutions like Midjourney will prove that statement false and show that AI can absolutely grow at a much faster rate than expected.
Now, getting back to blockchains and the integration of artificial intelligence, what really is the value here?
I like looking at the use cases that are often overlooked and there's probably a ton of them but today, I'll focus on one:
AI validators: securing the distribution and value of data onchain
Everybody knows one thing to be true about blockchain networks and that's that data posted to it is publicly available to anyone.
This raises a serious threat to the integrity of data as misinformation can spread rapidly and simultaneously creates a situation where bad players can attack individuals and institutions through actions like leaks or release of copyrighted materials publicly, etc.
Now I know that it's a common belief that blockchain, in its basic nature, ensures the integrity of data. The reality is that blockchain ensures that all information posted onchain remain as they were posted, the networks don't, in their current design validation the “accuracy or authenticity” of said data.
In essence, it's only half way there in ensuring the integrity of data made available to people.
With that misconception out of the way, let's explore how AI would help.
First off, we are talking about decentralized AI, so not only is the base settlement layer for all data decentralized, the pre-processing layer, with millions of AI models(agents), is also decentralized.
So, a use case for the synergy called “AI-blockchains” is that AI agents can become validators on the chain-level.
What exactly would they be validating? Generally, AI agents will be involved in 100% transactions happening onchain, but they will not be responsible for validating everything.
This means that whilst they will read all transactions coming through a network, they will not be responsible for validating all of it, so they'd only be working along stake-based validators to ensure the security of blockchains.
The type of transactions that AI will not be responsible for validating are purely monetary transactions. That said, they will primarily be responsible for validating transactions that include image, audio and video content.
Why? It's crucial that decentralized AI analyzes every audio, image and video content that get posted onchain to enforce the necessary copyright rules and also prevent the publicizing of illegal content such as child porn.
Artificial intelligence will be able to analyze these sets of data extensively and recommend(flag) against the inclusion of harmful data onchain. In the case of copyrighted materials, it really just depends on how the rules are implemented. For instance, if intellectual property owners welcome the idea of each use(posting in this case) of their content to include a $5-10 fee, then AI validators will be responsible for analysing if the initiator of said transaction has paid enough fees and included royalty beneficiary address that receives the % of the fees that covers for the legal terms.
In scenarios where enough fees are not paid and/or there's not the inclusion of appropriate addresses, then the transactions will not be added to the chain until the rules guiding contents included in it are met.
Certainly, these are just rough ideas that can be built upon into something really unique. The flaws associated with stolen content would be tackled and lost revenue would be recalled.
This is only going to be efficiently executed with the help of decentralized AI.
That said, above protecting copyrighted materials, ensuring the authenticity and accuracy of data posted onchain, be it text, image, audio or video will also require decentralized AI agents that can autonomously analyze and cross-reference with multiple on and off-chain sources, to derive a conclusion on the state of the data posted.
Understanding that there will be millions of models and all of these processes will likely be incentivized makes it easy to picture how effective it will be.
VCs may not necessarily have all of this mapped out as a reason to invest in this emerging synergy, but the value proposition remains evident as there are more ways than one that “AI-blockchains” can be game changing.