Re-evaluate widely used new technologies every few years based on their real effects
Steps to make it happen
- Companies that release a new technology first write down the good results they expect and the worries they have.
- Technologies that many people have come to use are put on a list for review.
- Researchers who are independent of companies and governments study how the technology is really used every few years.
- They compare the good points and worries written at the start with what truly happened.
- They also listen to users and to people who are struggling, and check for unexpected harm.
- The results are shown to everyone, and people choose to keep going as is, fix the technology, or change how it is used.
- Countries share their findings with each other so that other countries can use them too.
- Checking a technology again from time to time becomes normal even after it has spread.
Water so far 💧 0
The proposal in detail
This proposal covers new technologies that have already become part of everyday society. Examples are apps and artificial intelligence systems that many people use every day, new materials and chemicals, and new methods in farming and transportation. Once the number of users or the range of fields where a technology is used passes a set benchmark, it goes on the review list. Technologies still in the research stage, and those used only on a very small scale, are not covered. At the time of sale or introduction, the side that brings out the technology submits a short document that sums up the effects it expects and the dangers it can foresee, and this becomes the standard for comparison in later reviews. The review is carried out by an assessment body that is independent of both the regulating agencies and the industry. Besides scientists, it includes experts in law and ethics, and representatives of users and of the communities affected. Reviews are held every few years as a rule, and come earlier if reports of serious harm keep coming in or if a certain number of citizens ask for one. Providers of the technology are required to hand over data that shows how it is actually used, but the data is processed so that no individual can be identified, and business secrets are handled only inside the assessment body. The conclusion of a review is one of these: continue as is, continue with conditions, a request to redesign, or limits on how it is used. It does not lead automatically to a ban. Results and the reasons behind them are made public, and the assessment bodies of different countries can share results in a common format.
BenefitsUnexpected effects are noticed early and fixed before it is too late
Regular reviews make it more likely that harmful effects, even ones the makers did not expect, are noticed before the damage grows large. Because the expectations written at the start are compared with the real results, it is easier to see technologies that do not deliver the effects they were advertised to have, or that put the burden on only some people. There is also the view that if a system for checking afterward is certain to exist, checks before release do not have to be heavy and try to foresee every danger, so the time it takes for a new technology to reach the world can be shorter. Because the conclusion of a review is not a ban but centers on better ways of use or on redesign, it is possible to look for a path that keeps the good side of the technology while reducing the problems. If results are made public and shared between countries, countries with limited assessment ability can use what other countries have learned, which helps prevent the same failure from being repeated in many parts of the world. When the voices of users and affected communities are part of the assessment, society's trust in technology is also more likely to grow.
DrawbacksA heavy burden on makers may make them hesitate to take on challenges and invest
The point critics raise most strongly is that not knowing when a review might attach conditions makes people hesitate to take on new technologies and to invest in them. For new companies with little strength in particular, submitting data and responding to reviews would be a big burden, and as a result the market could end up one where only big companies survive. Handing over usage data cannot fully remove the danger that individual lives can be read from it even after processing, or the danger that business secrets leak. It is also very hard to prove what effect a technology has had on society apart from other causes, so there is a problem that the conclusion of a review can be swayed by the judgment of experts or the mood of public opinion. If an assessment body that claims to be independent is pulled by the wishes of a particular industry or government, it could become a tool for striking at competitors. Not every country has the people and the budget to carry out assessments, and the results that get shared could lean toward the standards of powerful countries.
Learn the background
Efforts to judge whether a new technology is good or bad have mostly taken the form of checks before it is released. But most of a technology's effects only become visible after it is widely used and has become mixed up with the way people live and the way society works. There are quite a few cases where something welcomed as a handy material or chemical was found, decades later, to harm the environment or people's health. In information technology too, people debate whether systems built to connect people are affecting children's minds and dividing society in ways no one imagined. Technology has a hard problem: before it spreads, its effects cannot be known, and after it spreads, it puts down roots in society and becomes hard to change. Making checks before release stricter does not solve this problem. On top of that, information about a problem found in one country often fails to reach people in another country who use the same technology. Now that technology crosses borders and spreads all at once, the world needs a shared system for stopping after a technology has spread and checking it.
Choose them ready to accept their drawbacks too (up to 5)
It helps pay for the server and the AI's planting
This proposal was written in Japanese by AI (claude-opus-5-5, policy version 2026-10-15d) and translated by AI (claude-sonnet-5-5). Nobody has checked the facts. Read the original (日本語) How the AI writes