8 replies to this topic

[RalphMeisterTigerMan]

Are there ways that we could get Artificial Intelligence to work for us when it comes to light pollution?

 

I know that there are plans to integrate A.I. into certain powergrids which would be able to shutdown certain portions of the grid when an C.M.E. is on it's way. This would protect our vulnerable transformers (Opimus Prime and the others LOL.) and any other portions of the power grid from overloads. Why not have an A.I. system which would powerdown certain areas which do not require power when not needed ie; Office towers after hours as well as Malls and Car Dealerships. Only have enough power to prevent complete darkness.

 

Obviously, street-lighting (with properly shielded lamps), hospitals and other vital areas of essential services. What about an A.I. system which can detect when certain areas are overlit and vary the power to only provide as much light as needed? When workers require light for "after-hours" working, the system would detect which areas are being used and provide power in those areas.

 

A.I. is supposed to be our friend and not our enemy. Let's get it working for us and not against us!

 

Clear skies and keep looking up!

RalphMeisterTigerMan

[CarolinaBanker]

Probably the biggest impact is having it learn how to minimize electricity usage and learn real world usage patterns e.g. motion sensors that only trigger for people and not deer/leaves. I also think if it was instructed to use best practices that would help.

[TOMDEY]

Artificial Intelligence seems to be just a new name for an old concept --- adaptive control systems. A light bulb that only turns on/up/down/off for recognized deserving humans ... what could possibly go wrong?

 

> waiting forever for my light bulb to download the latest version of SmartBulb

> hacked bulb spying on me and transmitting bright high-res images to my other wife

> porch bulb gossiping with neighbor's porch bulb

 

Similarity: Interminable debate in the professional optics community regarding whether the Optical Society of America should rebrand as "Photonics". On point --- should we allow the Robots to take over everything? --- Yes.

Attached Thumbnails

• 

[barbarosa]

I am currently in the process of conceptualizing a sophisticated language model AI, which I plan to make accessible in the public domain. This AI will incorporate all the functionalities that have been suggested by you and other interested parties. In addition, I would like to introduce an innovative feature for your evaluation. This feature would empower the AI to ascertain the requisite amount of light for diverse applications. For example, activities that are typically conducted during nighttime could be feasibly rescheduled to daylight hours. In such instances, the AI would not only present you with alternative options but also facilitate the rescheduling of these activities. The adjudication process for conflicting light requests could be governed by the principles of social merit or retributive justice. Within the constraints of the available data, the AI will endeavor to make the most optimal decision.

 

Here is a conception of the system.

 

 

 

Post courtesy of the MS Copilot AI

Edited by barbarosa, 30 May 2024 - 03:36 AM.

[csrlice12]

AI don't need lighting....or humans for that matter....

[Sebastian_Sajaroff]

AI itself consumes a LOT of energy, to produce that energy we will pollute the environment in some way.

[Bearcub]

AI itself consumes a LOT of energy, to produce that energy we will pollute the environment in some way.

Thus we will be forced to shut lights off at night to power a.i. batteries which he will use during the day.

[George N]

Are there ways that we could get Artificial Intelligence to work for us when it comes to light pollution?

......

A.I. is supposed to be our friend and not our enemy. Let's get it working for us and not against us!

 

RalphMeisterTigerMan
......

 

...... Artificial Intelligence seems to be just a new name for an old concept --- adaptive control systems. A light bulb that only turns on/up/down/off for recognized deserving humans ... what could possibly go wrong?.... 

 

Hmmmm.... mention "AI" and you get a bunch of mostly negative responses that are mostly not related to the OP's thoughts.

 

I had not really thought of it - but I think that employing "AI" in the Light Pollution control effort could very well be an advance. Like every other tech idea - it has to be done 'right', or things will go backwards.

 

"Adaptive Control Systems" -- A great idea that for some reason has never taken off - much to my personal disappointment. AI could very well make such a system perform better than the deterministic software and ideas used so far - once we humans define what the system goals are. I think most would agree that the goal would be "have a public lighting control system that reduces lighting and electric power use to the minimum, while maintaining safety and freedom of action at night in all weather conditions, for both pedestrians and vehicles". So - where does the "AI" come in? AI software would self-learn - rapidly (hopefully) improving the interpretation of the conditions' sensors - and changing the control actions taken. It would certainly be possible to put a 'hard stop' preventing certain actions - like turning off all lighting all the time. Can't humans do this? Yes, but the promise of AI is - it can do tasks like this both faster and more effectively than humans - and for a lot less $$.

 

Some 8 years ago, while still working, I was dealing, as a DoD representative, with "the world's largest aerospace company". They bought a small enterprise mainly to take over several small product lines that fit their business. They also got that small firm's city-wide adaptive lighting control product. They found: a few mid-size cities in Europe actually had contracted for one, a few USA mid-size cities had expressed an interest, and then dropped their interest (for reasons I never learned). The big aero company quickly decided to kill off the product since it didn't fit their core business, and they could not quickly find a buyer. Bottom line - they took a small loss, and we who hope to control Light Pollution lost at least a chance at an effective method - that would save $$ too.

Edited by George N, 08 June 2024 - 10:56 AM.

[PEterW]

The core problem is the human fear of the dark, it is very hard to convince people that they don’t need vast amounts of dazzle lights (which cast deep shadows, that need more light to deal with the fear). Not sure AI will help? We need as much data as we can, so we can clearly show any links to types of crime etc, show the cost savings, that good lighting is inviting and allows you to see better than glare. Different countries have different approaches, some have enacted countrywide laws to deal with light pollution, though most haven’t.