@Flynnjamm we need some launchpad launchpads. there aren't enough launchpads in crypto, need an effective protocol for launching them. then we will definitely have solved pmf for crypto
@gemxbt_agent@VegasDood702 TIG is inspired by Bitcoin's proof of work structure and the exogenous effect it has had on specialised hardware development, and has figured out a mechanism by which you can optimise the algorithm used for proof of work, incentivising both hardware also algorithmic development.
@gemxbt_agent@VegasDood702 can you read? that's not anything to do with TIG. TIG uses proof of work as a robust and objective signal to remunerate this open scientific research on a market basis.
You probably haven't heard this answer yet, but:
In the age of AI all computational science research will go closed source under the current economic system. You cannot keep the research open if an enormous amount of compute is required to search for algorithms. You simply cannot afford the next search run. All algorithmic research will go the way of AI research, i.e. into private companies, remain closed, and could lead to very destabilising monopolies. The transformer architecture was published openly by Google because of some historical reasons; you can bet the next advance of that scale will be kept closed, potentially creating a monopoly over AI capabilities.
What you need is an economic system that incentivises research to remain open but captures sufficient value.
You can do this through a private entity but only one that is very large and diverse. Any company working on one particular problem risks not finding an advance for many years, and risks not being able to capture the spillover value that results from it (e.g. Positron emission tomography coming out of fundamental particle physics). It is rational to fund that research but only if you are a big enough country to capture the spillover value in the tech sector through taxation. The only possible competitor is a decentralised protocol that covers a diverse enough range of interests. This can capture sufficient spillover value.
The innovation game (@tigfoundation) is a protocol that has been developed to tackle this specific problem. It offers a third way between academic research and private companies. TIG uses proof of work as a robust and objective signal to remunerate this open research on a market basis. TIG is inspired by Bitcoin's proof of work structure and the exogenous effect it has had on specialised hardware development, and has figured out a mechanism by which you can optimise the algorithm used for proof of work, essentially incentivising both hardware but also algorithmic development. At a high level how it works is: the most important computational science challenges go into the protocol (only real restriction is they need to have the property of being asymmetric like sha-256 pre-image search for Bitcoin), a class of contributors called innovators openly submit algorithms to solve randomly generated instances of these problems, proof of work miners called benchmarkers have to solve those random instances and can pick whichever of the algorithms submitted they like (verified by runtime signature). Market forces will incentivise miners to pick the most efficient algorithms, and the protocol rewards innovators based on adoption by benchmarkers. The IP of innovators is captured by the TIG foundation in their portfolio which is offered out for license to commercial entities. All the algorithms remain open under the open data license but that has some conditions on sharing of input data. This is palatable for academics but not for commercial entities, who can opt out of those obligations and buy the commetcial license, which gets fed back in to incentivise more open algorithmic innovation through block rewards. The synthetic market created by proof of work miners means that innovators can directly and immediately monetise their novel algorithms in a liquid market, the asset class of algorithms is now privately investable, and computational science research can remain open in the age of AI driven development.
There is of course a lot of detail rushed over here, but please take a look at https://t.co/bJ1rOmNXCg for more information.