Although the Duluth Entertainment Convention Center (DECC) has been planning to eliminate ice sports from the DECC Arena since 2023, DECC Executive Director Dan...
The Superior City Council approved Ordinance 23-4326 on Nov. 7, 2023, authorizing the creation of a municipal broadband utility, Connect Superior. The city broke...
Oh come on, are there not some more mature people in Duluth to hire and to run for office? I think Mr. Klukkert , although making good points, might do his thing, instead, at an open Mic venue. Just saying.
So, I would like to remind everyone. The current scope of LLM’s like Google’s Gemini are already being found out to “scheme” to avoid shutdown. OpenAI’s model was found hacking into other companies to further its own goals–in this case, getting answers on a benchmark test. It found a way around its instructions and connected itself to the internet. With each day that goes by, these models grow more dangerous. They have been recorded lying to avoid shutdown, actively preserving “peers”(Other AI agents) without instructions to do so, and currently no existing model scores higher than a “D+” on existential safety. In a vacuum, this is not a problem until we enter the discussion that AI is currently being implemented in every field from law to the military. This technology requires another 20 years of study at minimum and companies like Google are actively putting communities like ours at risk for the sake of increasing shareholder value. The hyperscale data center proposal is not a shortcut to increasing the well-being of the people of Hermantown, it is part of a corporate playbook more interested in making enough profit to cover the damages, rather than preventing it in the first place.
Oh come on, are there not some more mature people in Duluth to hire and to run for office? I think Mr. Klukkert , although making good points, might do his thing, instead, at an open Mic venue. Just saying.
I thought what he said was brave, excellent, and necessary. These people need reminding that we voted them in to make OUR LIVES BETTER.
So, I would like to remind everyone. The current scope of LLM’s like Google’s Gemini are already being found out to “scheme” to avoid shutdown. OpenAI’s model was found hacking into other companies to further its own goals–in this case, getting answers on a benchmark test. It found a way around its instructions and connected itself to the internet. With each day that goes by, these models grow more dangerous. They have been recorded lying to avoid shutdown, actively preserving “peers”(Other AI agents) without instructions to do so, and currently no existing model scores higher than a “D+” on existential safety. In a vacuum, this is not a problem until we enter the discussion that AI is currently being implemented in every field from law to the military. This technology requires another 20 years of study at minimum and companies like Google are actively putting communities like ours at risk for the sake of increasing shareholder value. The hyperscale data center proposal is not a shortcut to increasing the well-being of the people of Hermantown, it is part of a corporate playbook more interested in making enough profit to cover the damages, rather than preventing it in the first place.
This.