LLMs allow us to finally scale up our ancient chatbots (expert systems) and other fully artificial generation, analogous to original Moog. Why is no one intent on doing this?

Crazy.

0: The only possible form of democracy is a slightly-modified sortition system.

1: The "slightly-modified sortition system" must have a strict one-term limit, a strict lobbying ban at pension sanctiond, a some time of "attend 1/5th of lectures to pass" education for the winners, a blockchain for entropy storage & running whole lottery again as an app for mathematical proof of correctness, and smartphone restrictions for the winners.

2: It is very likely, that by forming a Random Party and getting it admitted as a legal "identity" of the candidates in the lottery (it is more of identity as small chance to chosen), such system could become part of the landscape in some of the economically smaller and less significant to nation's strategic nuclear deterrence in Europe. With a Random Party existing, and votable, there would be no need to immediately gain constitutionally-affective parliamental control required for a staying reform.

3: The motivation besides removing long political "careers", next elections pressure, influence of campaign ad spending, and many other diseases of politics obviously is improving economy for all. It is not only very likely that such mechanical permit would be juridically feasible, but also very likely that between 6% and 36% of the seats the party would begin to drastically eliminate market inefficiencies such as regulatory capture. Because the members of the Random Party have no re-election or donor-driven incentives, only their public image and post-term quality of life, their views will approximate those of the actual population. The added noise "brings the median voter theorem" to the (electoratial) parliament.

4: A website with strong European identity verification and the blockchain can be made with relatively low cost to expected improbable but large impact.

5: Things to do that affect success but are not "needed" would be proper mathematical and sociological and legal study to fill up all holes before any moles can raise (afterwards changing a system is difdicult so formal proofs and study to leave "nothing" unconsidered is nexessary), public education program, behavioural data using marketing and emergency PR team, a face that can do the inevitable hearings and interviews without ruining the project, and small group of financially and bureaucratically literate bookkeepees to handle labor of being a "party".

6: The blockchain part is not related to cryptocurrencies but ensuring that entropy comes from many sources, potentially even "anyone who wants to insert their own bits" in addition to institutional entropy sources, that the whole of entropy remains immutable such entropy pushes, that it and the candidate registrar as well as the source code is available for download in a manner resistant to low-to-medium difficulty censorship attempts, and critically that the hashing function can be rerun on consumer decices from source. It should also make use of for history proofs, however, the Open TimeStamps (OTS) project, which is based on Bitcoin.

7: All of this is both possible and sane to implement, and the when is likely a question of time. Almost none of this is original work except for the "how to actually execute it" planning.  It is very probable that if a small nation implements a sortition system such as described above, that conditional to them not getting immediately sabotaged, their economy will eventually begin benefit from sortition compared to other nations.

2025-06-18T18:59:58.000Z

0: I divide my treatment of intelligences into game theoretical examination of the uses of intelligences and EA theoretical examination of rationality in fictional envrionments ultimately aiming for a powerset of strategies for synthetic intelligences of different sizes.

1: I hold deep interest in computational ultrafinitism, i.e. limiting the sum of bits every automata state delta takes when added to the beginning state such that it never exceeds M bits including the limiting "kernel", which I see as the only way to pave mathematics forwards.