1,299 karma · joined January 5, 2013
http://ets.aeso.ca/ets_web/ip/Market/Reports/CSDReportServle...
Low unemployment? That's a labour shortage.
High GDP growth? The economy is overheating.
High rates of foreign investment? We are losing control of our economy.
Country experiencing historic levels of prosperity? Wealth is making everyone lazy, fat and unproductive.
Is this better?
Output in the US has been flat for some time, while China has been on a steady rate of climb for several decades.
Late 1990s internet infrastructure buildout means that most of the captured value was at the hardware level (HP Computers, Dell, Intel, Cisco etc.).
2020s AI infrastructure buildout appears to be the same. NVIDA is capturing a lot of value, and experiencing explosive growth.
Late 1990s/early 2000s was a time of tremendous development for web software , but no one really had a business model. These LLMs are like the web browsers of that era -- they are an interface to access other peoples IP, but no one really knows what the financial model looks like.
It wasn't until the mid 2000s/early 2010s that the business models came into sharp focus for web services.
Like web browsers, the LLMs might be a vital tool for enabling the capture of value for businesses, but it might not be the actual thing that they end up monetizing.
I can think of a million different software services that have some value to users, but don't have some multi-trillion dollar revenue stream flowing from them.
There is an idea that these LLM companies are going to be able to insert their agents into the labour market and capture some percentage of the savings that firms realize from laying off their human workforce. Given the competitive environment, it is far more likely that these LLMs become an undifferentiated commodity good, and their value to the economy gets captured elsewhere. Currently the only value being captured is at the infrastructure level, and that is all predicated on a series of assumptions around software business models that have not materialized yet.
The black box algorithms are the problem.
Where laws and regulations get complicated is when they are referring to other documents. The Canadian Parliament has done a wonderful job creating a user interface for the different iterations of bills, and linking them to video and text of the debates and votes. https://www.parl.ca/legisinfo/en/bills
Facebook seems full of older people interacting with AI generated visual content who don't seem to understand that it is fake.
Our society already had a problem with people (not) participating in consensus reality. This is going to pour gasoline on the fire.
There are many jurisdictions that can afford to do both, but most are not in that position.
There is far more wasted potential in the case of the hundreds of thousands of kids who fall through the cracks because of a crummy education system. Many states already create an uneven playing field by funding each school system based on the quantum of local taxes collected in their particular communities. Poor kids shouldn't have to further compete for resources within their own, poorly funded institution.
In a properly functioning team, people perform different, discrete roles which are probably not entirely understood by other team members or management.
- A lot of companies shut down for the period between Christmas and New Years. That should be a check box.
- There should be user preference for vacation during warm or cold weather seasons.