From a value perspective, the main difference between foundational models and crypto tokens is that foundational models can be considered a type of public good, meaning that they can be easily replicated and used by anyone without restriction. This is the case for models like GPT-2, which have been made publicly available by OpenAI. In contrast, a crypto token is a private good that can only be used by a single person at a time and serves no other purpose than as a store of value.
While foundational models can quickly become obsolete as new and improved versions are developed, crypto tokens tend to retain their value over time. This is particularly true for coins like Bitcoin, which have underlying appreciation built into their value.
When a company like Microsoft invests a large sum of money, such as $1 billion, into a company like OpenAI, that investment is used to create a new and improved model like GPT-3 or ChatGPT. This type of investment creates an asset that can benefit society as a whole, as it can be used for a variety of applications and can be shared and used by anyone. In contrast, if that same investment was used to mine crypto tokens, it is less clear how much additional benefit would be created for society as a whole. Overall, investing in foundational models can create a greater social benefit than investing in crypto tokens.
The text above is a ChatGPT rewrite of this human original:
"From a value perspective, the difference is that foundational models can potentially be public goods (like GPT-2) that can be copied again and again and used by everyone, whereas a crypto token is a private good that explicitly cannot be used by more than one person (and doesn't do anything except exist). Foundational models, however, do quickly become less useful over time as they become obsolete, but coins remain more-or-less constant value (with underlying appreciation usually built in, like with BTC). When Microsoft invests $1 billion in OpenAI which uses that to melt electricity and silicon into GPT-3 or ChatGPT they create an asset that all society can benefit from (a true Public Good. If they had spent that same money to mine some coins it is hard to see how much incremental social benefit would have been created."
Image created with Dall.E, (c) Lawrence W. Sinclair
Monitoring misinformation and disinformation in U.S. Spanish language radio broadcasts using AI.
Computer use refuses to act as an agent representing people.
“sign up for the next available workshop. At...
Foundational models are Public Goods. The real money for investors will be in their applications.
Google's experimental Notebooklm produces a podcast of your docs.
The biggest factors contributing to companies’ bottom lines right now are high interest rates and artificial intelligence, says...
A similar approach to prompting to create a chat app like ChatGPT.
Large Language Model (LLM) developers such as OpenAI have no concern about open sourcing their algorithms.
AI will not put people out of work, at least not in the long run. But our roles...
Many outspoken people who write frequently have criticized ChatGPT and similar language models as crutches, claiming their users...
Efficient storage and retrieval of high-dimensional vectors are critical for maintaining performance and scalability. The halfvec data type...
