Perspective by: Merav Ozair, PhD
Tech leaders continuously celebrate the artificial intelligence transformation — from Bill Gates to Sundar Pichai to Jensen Huang — indicating that autonomous AI and robotics will take over our jobs and serve as our self-sufficient aides operating on our behalf both in professional and personal realms.
Regardless of whether these projections materialize in a few years or stretch into decades, we will most likely transition into this future in some form, and technology, yet again, will redefine our existence. However, without the backing of blockchain technology, it would be challenging, if not impossible, for autonomous AI and robotics to advance to the level their advocates envision.
If we desire these services and tools to operate independently, considerations surrounding security, privacy, transparency, and accountability will dominate our thinking. These are the domains where blockchain excels and can mitigate the deficiencies of AI to enable the expansion and development of this vision.
Blockchain advantages bolster AI limitations
Blockchain technology can substantially enhance the security of AI models by utilizing its essential characteristics, including decentralization, immutability, traceability, smart contracts, data privacy, and identity confirmation. For instance, but not limited to:
-
The decentralization feature removes a central point of vulnerability, augmenting the robustness of AI models against infringements.
-
The immutability of blockchain guarantees that the data utilized in training AI models and the models themselves cannot be unlawfully modified, preserving the integrity of the systems.
-
Every change or action taken by the AI model can be audibly tracked through blockchain, offering unmatched transparency and accountability.
-
Smart contracts automate the enforcement of data access and usage policies, preventing unauthorized or unethical exploitation of AI models.
-
Smart contracts can guarantee that data is utilized solely for training and testing by authorized individuals, locking the possibility of being applied for other uses. Merging these regulations with multiparty computation could thwart or at least lessen AI adversarial assaults.
-
Blockchain facilitates secure multiparty computation, ensuring data confidentiality during AI model training by keeping the information decentralized.
-
Blockchain’s secure identity verification enhances the security of AI systems by thwarting unauthorized access.
Combining AI with blockchain can create a safe, transparent, traceable, and decentralized AI ecosystem, safeguarding our privacy, boosting accountability, and realizing responsible AI.
Transactions: Programmable AI intersects with programmable blockchain
AI agents and robotics are adaptable. Smart contracts, the backbone of digital assets, are also programmable. It is entirely logical that digital assets would serve as the preferred payment pathway for agent-to-human and agent-to-agent interactions, which encompass robotics.
Cryptocurrency is an internet-native, programmable currency with numerous benefits for fueling the agent-based economy. As AI agents gain more independence and engage in micro-transactions at scale, crypto’s efficiency, global accessibility, and programmability will render it the favored medium of exchange compared to conventional fiat systems.
Recent: Sentient open-source AI search surpasses GPT-4o and Perplexity
The real convergence of Web3 and autonomous AI for financial dealings could materialize through new tokens and protocols designed for this purpose. These could expand stablecoin functionalities by incorporating agent-specific elements.
In this scenario, payments could occur using a specialized asset that agents can stake for quality assurance. Slashing policies could penalize inadequate performance, while validators could handle disputes based on task quality.
Moreover, agents’ reputations could be directly connected to their token stakes. Implementing rules through smart contracts empowers users to manage their autonomous workers/assistants, allowing for a shutdown or even a “kill switch,” if necessary, during instances when AI agents begin to act recklessly.
If Goldman Sachs wishes to develop AI agents that think and operate like a seasoned employee in a highly regulated sector with critical risks to financial systems and the overall stability of financial markets, it would be essential, not optional, for these AI agents to be governed by programmable tokens.
Although this strategy necessitates advancements in both Web3 and autonomous AI, it is not as far off as it may appear.
Blockchain development organization Skyfire recently introduced a payment network that allows AI agents to spend money autonomously. Led by former Ripple vice president of products and services Amir Sarhangi, the company’s platform enables a business to provide a pre-loaded wallet to an AI agent.
The firm’s protocol converts the cash into USDC (USDC). In early March, Skyfire launched its payment network enabling AI agents to execute autonomous transactions out of beta.
Utilizing digital assets for robotics, VR devices, and autonomous AI transactions transcends a mere payment method for exchanges. It could vastly improve user experience and security while creating countless novel business models that have yet to emerge.
It will be fascinating to observe how this all unfolds and if other companies will follow suit.
Nonetheless, there are risk factors that need addressing, and we must be conscious of how they are, at the very least, mitigated. This is where we should carefully contemplate the security protocols mentioned earlier.
Moving beyond “tunnel vision” to a multifaceted strategy
There exists a significant emphasis on the progression of AI — generative AI, autonomous AI, reasoning models, physical world models, among others — all centered around the notion that AI is the singular technology required to develop AI autonomous agents on a large scale.
This represents a narrow viewpoint in how products are developed and reflects a certain shortsightedness: neglecting to recognize what must be achieved beyond the enhancement of AI models for the ecosystem to develop and proliferate.
AI, no matter how advanced, cannot exist in isolation and requires the support of blockchain technology — a programmable connection ideally suited for complementary development. Thus, we must adopt a multifaceted approach. We should consider and treat AI and Web3 together regarding innovation, regulation, and infrastructure. This is pivotal to the foundation of a thriving autonomous economy.
“Dreams are constructed on solid foundations,” and the time to construct them is now.
Perspective by: Merav Ozair, PhD.
This article serves general informational purposes and is not intended to be nor should it be construed as legal or investment advice. The views, thoughts, and opinions expressed herein are solely those of the author and do not necessarily reflect or represent the views and opinions of Cointelegraph.