Disclosure: The views and opinions expressed herein belong solely to the authors and do not represent the views and opinions of crypto.news editorials.
The growth rate of AI is outpacing regulation, creating risks to data, identity, and reputation verification, and if left unchecked, could increase the prevalence of misinformation and slow progress in scientific innovation. The push toward superintelligent AI is described by some of its most passionate leaders as a push toward a golden age of science. However, this push poses an existential risk that our society will reach a ceiling on large-scale adoption of immature AI technologies and enter a state of degrading technology stagnation, where human creativity and innovation diminish over time. is likely to increase.
This is a contradictory view for most accelerationists. AI is thought to improve our ability to perform our jobs faster and synthesize larger amounts of information. However, AI cannot replace inductive reasoning or experimental processes. Today, anyone can use AI to formulate scientific hypotheses and use them as input to write scientific papers. Results from products like Aithor often appear authoritative on the surface and may even pass peer review. This is a big problem because AI-generated texts are already cherry-picked as legitimate scientific findings and often contain fake and fabricated data to support their claims. There is a strong incentive for young researchers to compete for a limited number of academic jobs and funding opportunities using all available means. The current incentive system in academia rewards those who publish the most papers, regardless of whether the papers describe valid findings. All you need to do is pass peer review and get enough citations.
Academic content with unconfirmed authors poses a serious problem for industries that rely on basic science to promote research and development. It will maintain the quality of life. As a result, well-funded R&D can only rely on research that it can perform and reproduce, increasing the value of trade secrets and dealing a devastating blow to open science and access to meaningful information. That’s it.
Expensive replication efforts alone can address misinformation, but the problem is much bigger than that. Today, we face a loss of trust in the very foundations of knowledge, with unverifiable claims and ambiguous attributions undermining scientific progress and threatening the scientific community. There is an urgent need to establish a truth-based economy that reliably authenticates content and data.
AI systems are only as powerful as the data they are trained on
Large-scale language models are great tools for generating persuasive content. However, they only provide as much information as the data they are trained on. The ability to extrapolate outside of the training set is still limited. The role of science is not only to synthesize existing knowledge, but also to create new and useful artifacts that increase the entropy of humanity’s collective body of knowledge. Over time, as more people use AI to generate content and fewer people generate original content, we will not be introducing new information into the world, but simply combining past knowledge. You will be faced with “low entropy bloat”. Unless we build resilient provenance and verified attribution layers into AI tools used for serious research, new “knowledge” will be based on AI-generated self-referential content and primary sources will disappear. I will.
This “lobotomization” of the intellectual depth of the human population will have lasting effects not only on art and creative activity, but also on medicine, economics, and academic research. Unverified data can influence research, skew results, and lead to important policy and technology failures that undermine the authority of scientific research. The risks of “science” created by AI are wide-ranging. The day-to-day operations of normal science will be bogged down by disputes over authorship, allegations of plagiarism, and obstacles to peer review. More time and energy will need to be spent to address the many consequences of declining quality and precision in scientific research.
AI is a useful tool for generating ideas, structuring thinking, and automating repetitive tasks. It must continue to complement human-generated content, not replace it. It should not be used to write scientific papers proposing original discoveries without doing the work, but as an aid to increase the efficiency and accuracy of human-driven efforts. For example, AI can help run simulations on existing data in known ways and automate this work to discover new research directions. However, the experimental protocols and human creativity required for scientific investigation cannot be easily replaced.
Building an economy based on truth
A truth-based economy establishes a framework with systems and standards to ensure the authenticity, integrity, transparency, and traceability of information and data. This addresses the need to establish trust and verifiability across the technological world, allowing individuals and organizations to rely on the accuracy of shared knowledge. Value is rooted in the truth of claims and the reliability of observations and primary sources. A truth-based economy will make digital knowledge “hard” in the same way that Bitcoin made fiat money hard. This is the promise of the decentralized science movement.
How do we get there? We need to start with the most important element in the world of science: the individual researcher and his or her work. Today’s current web standards for scientific identity are insufficient to verify identity claims and proof of work. Current practices make it very easy to create a profile with a decent reputation. Peer review is also compromised by bias and collusion. Without validation of the metadata that accompanies scientific claims, we cannot establish a truth-based economy for science.
Improving academic identity standards starts with simple cross-platform logins powered by privacy-protecting identity technology. Users should be able to sign in to any site using their credentials, prove their trustworthiness, and selectively disclose reputation, data, or facts about other agents or users.
An identity layer rooted in verifiable researcher reputation is the basic foundation of DeSci. A fully on-chain scientific economy enables anonymous participation of the public in large-scale online coordination for research activities. Research institutions and decentralized autonomous organizations can create permissionless systems and bounty programs that are not fooled by fraudulent reputation or identity claims. A universal scientific registry secured on a blockchain with identity claims would provide a frame of reference for autonomous organizations built to accumulate verifiable scientific knowledge and test falsifiable hypotheses. Masu.
Protecting the future of human progress
To avoid a collapse of trust in professional research fields, we need to establish a foundation of truth through transparency and rigorous verification of information. The potential for our collective progress to continue for hundreds of more years and unleash successive scientific revolutions in materials science, biotechnology, neuroscience, and complexity science depends on high-quality research and curation of sound data . This will be the difference between a future society as advanced as ours and a pre-Enlightenment society. Otherwise, we have to hope that this is as smart as it can be as a species, and we’ll only get dumber and dumber. It’s not clear that DeSci will save us, but time is running out to make things right.
