Author: Muhammad Yusuf, Delphi Digital Researcher

Translation: Yuliya, PANews

Editor's note: Although BIO has experienced a surge over nearly a month, do not be misled by the short-term wealth myths and token frenzy. When AI and open-source models are lowering the bar for new drug development to freezing point, genuine medical breakthroughs still cannot bypass the long and rigorous clinical cycles. In the face of the crypto circle's impatience for speculative hype and the "funding winter" of traditional research, DeSci has indeed welcomed a breakthrough opportunity. However, what it requires is absolutely not a speculative game that changes tracks in just a few months, but needs to establish a financing infrastructure strictly tied to clinical milestones, governed by professionals, to ensure sustainable development.

An AI Agent named peptAI designed a new ADHD (Attention Deficit Hyperactivity Disorder) peptide drug candidate from scratch in just 24 hours, passing through eight validation processes, and outputted a molecule ready for direct wet lab testing. The laboratory work cost only a few thousand dollars. As a supportive platform for this project, BIO Protocol's token skyrocketed by 105%. Within just a few hours, half of the personal profiles on Crypto Twitter added the term “DeSci,” just as they did six months ago when they added “AI.”

Now, open-source protein folding models can rival AlphaFold3 at zero licensing cost, public databases of bioactive compounds cover 2.5 million compounds, and the verification costs in wet labs have dropped below $2000. AI is significantly compressing the costs and time of drug development. Over the past week, I have been trying to figure out what is truly different this time.

Surviving Phase 1 Clinical Trials Means Nothing

Current circulating data suggests that drugs discovered by AI have a success rate of 80-90% in Phase 1 clinical trials, while the traditional baseline is only about 47%. However, no one adds that Phase 1 clinical trials test whether a drug can cause death, not whether it can cure a disease. Getting through this stage only means that your compound is safe enough to continue research, but it still needs to undergo layers of screening until it ultimately obtains FDA approval.

Currently, fewer than 40 AI-discovered compounds have reported Phase 2 clinical data, and none have completed Phase 3 trials. Insilico Medicine's Rentosertib is currently the most advanced AI-discovered compound, which announced positive Phase 2a results for idiopathic pulmonary fibrosis in mid-2025 (published in Nature Medicine) and began Phase 3 clinical enrollment in China in the fourth quarter of 2025. If all goes smoothly (with 2027 for enrollment completion, 2028 for data readout, and 2029 for FDA review), for the best candidate among this pipeline of 173 drugs, you will need to wait at least three years. Several compounds in this pipeline were put on hold in 2025 for failing to meet primary endpoint criteria for atopic dermatitis, schizophrenia, and cancer. Independent analysts estimate that the probability of the first AI-designed drug receiving FDA approval in 2027 is 60%, but so far, no AI-designed drug has completed this process.

Is Crypto Twitter Really Suitable for Genuine DeSci?

Now, keep these timelines in mind, and take a look at the chart from BIO Protocol. The token fell from $0.89 to $0.018, then rebounded 105% stimulated by the peptAI news, with a trading volume of $720 million, and a market cap of approximately $68 million. The entire financing logic of DeSci is based on one assumption: that token holders will patiently wait for clinical projects lasting seven to ten years, yet the reality is that even before Phase 1 clinical data is unblinded, Crypto Twitter has already turned to the next narrative.

Pump Science even leaked its private key on GitHub, spawning a plethora of fraudulent tokens, one of which is directly called Cocaine, and the enforceability of IP-NFT has never been tested in court.

The Contest Between Open Source and DeSci

If we do not fall into the reflexive self-deception and long-term speculative hype, open-source science might bring a glimmer of hope to DeSci.

In October 2025, the OpenFold Alliance released OpenFold3 under the Apache 2.0 license. It is fully trainable and commercially usable, based on structures determined from over 300,000 experiments (unlike AlphaFold3 which is restricted by Google for academic use only). Boltz-2 from MIT and Recursion predicts protein structures and binding affinities 1000 times faster than physical-based methods.

The Baker Lab released RFdiffusion3 in December. ChEMBL hosts 2.5 million bioactive compounds with complete ADMET distribution, available for free to anyone with a laptop. In the past, pharmaceutical companies had to spend millions of dollars building infrastructure in-house; now it’s available on GitHub under flexible licenses, and five pharmaceutical companies are currently federating training on their proprietary drug-protein libraries through the Federated OpenFold3 Initiative.

No one on Crypto Twitter talks about these, as there are no tokens to trade, and I highly doubt whether the core contributors of these codebases would be excited about issuing tokens.

Funding Shortage in the Scientific Community

In 2025, over 7,800 NIH and NSF grants were terminated or suspended, with over $5 billion in funds frozen. NIH (the largest public funder of biomedical research globally, with an annual budget of approximately $47 billion) has kept its budget unchanged as Congress continues to allocate funds, but the government has still frozen the funding pipeline. The number of new competitive grant awards dropped sharply from 11,659 in fiscal year 2024 to 6,095 in fiscal year 2025, a decline of 48%. The success rate of researchers applying for funding fell from 21% to 13%, and Fred Hutch lost $508 million, while Harvard University lost $945 million.

This funding gap is exactly why DeSci has the opportunity to sell itself if done correctly. In July 2025, Gero, funded by VitaDAO, signed a research and licensing agreement with Chugai Pharmaceutical (a Roche subsidiary valued around $100 billion), with milestone payments of up to $250 million. This is the first proof that a DAO-funded project can produce something valued in the nine figures by a real pharmaceutical company. This process successfully transpired without governance struggles or project collapses, remaining one of the most significant events in the field.

A Long Four Years

This year, 15 to 20 AI-discovered drugs will enter Phase 3 trials, while the data readout for Rentosertib will not occur until at least 2028, meaning it will take several more years to reach a definitive conclusion on whether all of this can translate into drugs that work in humans. Regardless of the existence of tokens, the open-source tech stack will continue to compress costs, while the funding vacuum will continue to push researchers toward anyone willing to write a check. Today, open-source protein folding models can rival AlphaFold3 at zero licensing costs, with wet lab verification costs dropping below $2000, while the NIH has just announced the lowest grant success rate in twenty years.

Even if AI fulfills all its supporters' promises and compresses the drug development timeline by half, from discovery to approval still requires a cycle of four to five years, and this is assuming that the success rate for Phase 3 clinical trials indeed improves in an optimistic scenario. In an industry where portfolio beliefs rotate with quarterly earnings calls, four years is a long wait; token holders even consider holding for six months as a type of indefinite sentence.

The costs of drug discovery and innovation are decreasing each quarter, regardless of the popularity of tokens. The funding shortage from the NIH is also concerning. Perhaps between these two, there is a viable model where tokens can fund specific trials and milestones, while governance is managed by professionals. The collaboration between Gero and Chugai is the first proof that DAO-funded projects can produce something that pharmaceutical companies are willing to pay nine-figure sums for. Besides eliminating the hype, I am also curious if anyone will build a defensive financing infrastructure for genuine decentralized science.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。