taotau
Joined 6 years ago
Joined 6 years ago
Sure an LLM could write it's own libraries and abstractions in a low level language, and im sure there are some assembler or c level web api wrappers, but they would be nowhere near as comprehensive or battle tested as the ones available for high level languages.
This could definitely change in the future. I think we need a coding platform that is designed for optimised LLM use, but that still allows humans to understand and write it. Kind of a markdown for code. Sort of like what OP is trying to do, but with the built in benefit of having a common shared suite of tools for interoperability.
The only real advantage I had over the current crop of LLMs was the ability to reliably retain context between sessions, but even that wasnt very useful initially as every system was so bespoke.
I then moved on to using some level of social engineering to extend my ability to gain access to and learn about these systems.
Doing this over and over, I like to think I have developed some pretty complex understanding and abilities.
I ran a local NNTP relay on my home phone back in the day. I recall spending quite a bit of time filtering out a lot of the binary groups.
When data collection and manufacturing supply chains were more crude, it was easier to make blue and pink gi joes and barbies.
Now we have the abilitiy to collect more nuanced data about peoples preferences and efficiently service the long tail, it makes sense to let customers define their likes on a spectrum rather than as binary choices.
I dont see why you would feel so strongly about restricting how people choose to define themselves.
I would imagine that a planetary system as a whole has a 'climate' driven by it's host star, and redirecting large amounts of that energy would have unpredictable consequences. While most of the suns energy does leak off into interstellar space, before it reaches that point it interacts with various bodies big and small, solid and gaseous, it generates magnetic fields and powers phenomena we may not be aware of.
Perhaps the choice of creating a Dyson swarm IS the great filter, and any civilization that has achieved it finds itself in a state of 'solar climate crisis'.
While on the surface, the proposals sound very efficient and low impact, as always the devil is in the details.
Standard operation would use closed loop cooling, relying on rooftop rain water to keep the loops topped up. However, the company has acknowledged that when ambient temperatures rise above a certain level (26C), they would need to switch to evaporative cooling.
The contract with the government also has a section where the company requires a connection to the state agricultural water supply as a backup [1].
When rainfall is plentiful and temperatures are low, as they often are in tasmania, everything is rosy.
However the state does experience periods of draught and thats where things get sketchy. Conflicts are likely to arise when a large investor in the states economy wants to start draining the local water supply to keep ai agents running at a time when crops and livestock are competing for the same resources.
Similar issues would arise regarding power. Firmus is proposing to underwrite an expansion of the states already extensive Hydro power network, but for that investment they would very likely expect priority access to the grid when supplies start running low.
[1] https://www.tasmaniancountry.com/newspaper/tasmanian-country...