We are entering a new era when the ability to use and work with artificial and synthetic intelligence is a human right.
Access to intelligence – prerogative to innovate, work with and benefit from higher levels of synthetic intelligence – belongs to the people.
Based on increasingly cheap calculations, abundant data and low-cost open-source models, we are about to see a horn of abundance of synthetic intelligence.
We must build infrastructure that supports the pluralist development of AI. This is why we start the Thames network, based in Oxford: a decentralized intelligent network to operate at the forefront, allowing private AI, resistant to censorship, depoliticized and decentralized by integrated economic incentives and cryptographic evidence.
“I have concerns about the concentration of power and the loss of privacy to which AI leads. It is essential for us to think of strong technical solutions on this subject, such as blockchain, “said the principal scientist, the professor of Oxford Philip Torr, calling for incessant progress in a decentralized AI, putting the power of Ia to the people.
Decentralize the AI
What is this open source, the decentralized market, the protocol and the incentive layer for artificial and synthetic intelligence imply? The release of Deepseek has multiple implications, the most important of which is that open source AI is there to stay, and that the future made does not belong to a large centralized, corporate (or state) model. The AI has passed from the center to the edge, and it now becomes more decentralized.
Microsoft has just announced that distilled and optimized versions of the DEEPSEEK R1 NPU will be available on PCS, first taking advantage of local treatment, starting with Qualcomm Snapdragon X, followed by Intel Core Ultra 200V and others. Users will be able to interact with the new family of revolutionary models entirely locally.
Neuronal processing units are specialized computer microprocessors designed to imitate the process of processing the neural network of the human brain. The NPUs, which will be presented on personal devices, offer a very effective set of capacities for the inference of the model, unlocking the agency paradigm where a generative AI can execute not only when it is directly invoked, but allow operating services Semi and entirely entirely continuous, that is to say agents.
The movement towards decentralization is more than a technical upgrade. It represents a fundamental change in the way we empower individuals. This means promoting AI systems oriented towards collaboration, resulting in innovation while protecting itself against the traps of centralized control, explains Richard Sutton, widely recognized as the “father of strengthening learning”. “The learning of strengthening, rather than important language models, holds the key to the progress of the AI,” he said.
Democratize AI
The Time network democratizes access to AI with the first the Open Source decentralized AI market, the protocol and the incentive layer.
Universal basic income – where citizens are offered recurring payments to subsidize their lives – is presented by AI oligarchs, and in particular Elon Musk, if necessary. It is not an approach first of people; This is an approach first in a business, and which will eat out the fabric of society. The best approach is to democratize access to AI and allow autonomy and sovereignty for the individual.
With a new cutting -edge intelligence substrate and with a new economic model, a decentralized intelligent network would light up an ecosystem of agents working in concert with humans. Rather than subsumer or replacing humans, this network will create new opportunities to democratize human-AI collaboration.
Depolization of AI
For artificial and synthetic intelligence in favor of humanity, it is imperative that it is exempt from biases and to be apolitical, without an implicit (or explicit) program. Censorship, railings and access limitations based on jurisdiction, price and other factors are not the means to create a future where humans and AI can collaborate effectively.
At the same time, confidentiality is the key in areas such as health care. A decentralized intelligent network must be designed with an approach before confidentiality, and architects on a confidence basis, guaranteeing a zero-frust security model, while balancing governance, risk and compliance.
This can start with hundreds of thousands of models will meet in a massive wave of hundreds of millions of models specific to the field, organized, distilled and increased via a generation (RAG) (RAG) of recovery. The Thames network will provide tools and a market open to people construction and to monetize their domain expertise, once again by emphasizing human-Ai collaboration.
“We all see the digital world taking control of our world via the Internet, the collection and sharing of data and the current power of AI,” explains Bill Roscoe, director of the Oxford Blockchain Research Center. “The world really needs an altruistic development of the rules of digital civilization and an infrastructure to support and govern it in a truly collective way.” The mission of the Thames network is to ensure that privacy and collective governance remain at the forefront of technological development.
The convergence of decentralized IT, tools and governance of blockchain, protocols and mechanisms of cryptographic incentive and models of AI specific to the field built and organized by human experts, indicates a future where intelligence Artificial and synthetic becomes accessible, transparent, secure, abundant and collaborative.
The Thames network, which we are announcing at the Oxford Ai X Blockchain conference today, plans a win-win world for humans and AI. Anything that would be an abdication of responsibility as technologists, engineers, researchers and economists.
