Community Info, Safety and Support
Get informed. Promote safety, fairness and preparedness for all.
A SECURE AND RESPONSIBLE APPROACH TO A.I.
HUMAN EXTINCTION LEVEL????
According to Brookings.edu:
The human extinction concern is a future risk scenario, NOT a prediction that it will happen. Human extinction level can only happen if advanced AI bot agents were given bad goals and too much "leeway" to act—through software, infrastructure, or other tools—it might pursue those goals in harmful ways. So, “Stopping the AI” might sound fearful, but it usually means submitting proposals to slow or pause the development of Advanced A.I. systems called Frontier AI models also known as A.I. Agent Bots placed in a system to set or follow specific goals. "Stopping the AI" does not mean turning off every A.I. tool. BUT A.I. COMPANIES HOLD THE POWER TO CONTROL OR CONSTRAIN THESE BOTS.
But because these bots can be highly capable of carrying out important tasks by the user, if these bots are designed and built and then told to do the wrong thing by the user, or it mistakes a word or phrase for something else, it could make grave mistakes. So, if misused or not checked before allowing these bots to be released to take action, these bots could become "rogue" and cause serious harm. For example, an AI-generated deepfake impersonated company executives and coworkers, persuading them to make a large money transfer enabling financial fraud. Researchers are currently working on making these A.I. bots follow intended instructions to keep them contained and under control:
OpenAI, Anthropic, Google, DeepMind, Meta and research labs are currently working to circumvent these harms because again, they hold the power.
SUGGESTED TIPS TO STOP ROGUE ADVANCED A.I. BOTS OR ROGUE BEHAVIOR
1.Bipartisan Collaboration is needed across local, city, state and national governments--especially here in the United States as well as international collaboration.
2. Monitor and Stop the Public Data Breaches. Representatives and Senators in Congress, Stop the Public Data Breaches. This is where the bot misusage started. The public has experienced several data breaches without monetary compensation for the invasion of their privacy. Dates and names should have be kept on WHO experienced the Data Breaches and WHEN the Data Breaches occurred. Many people in the public have been unknowingly sent viruses or codes attached to their emails or other personal information and when sent unknowingly, goes on to hack or attack larger corporations "as replicators" each time their email or other information is sent. PAY ATTENTION TO DATA BREACHES THAT ATTACK THE PUBLIC.
3. Create A.I. Apps and surveillance tools and techniques to flag or locate listening devices to warn the public, A.I. developer and business design locations or high-profile corporate meetings while in private settings or business settings in the United States.
4. Limit what A.I. can help with, make bots ask before making important decisions about a person or taking important actions to reach a goal; test the A.I. carefully before deploying, and keep an eye on WHO is making the A.I. chips, WHO is using it and HOW it’s being used.
5. Level and Label Tasks. Keep Dates.
CIVIL:
Everyday use--tasks that people use
economical use--tasks that businesses and organizations use for the economy
Corporate use--tasks that large corporations use
More...
(CONTINUE)
City, State Laws--tasks for law officials
Everyday law officials
Investigative law Officials
Courts--lawsuit tasks
Medical use--tasks that medical personnel use; tasks for surgical use
Law officials--
Court Officials-
Military
U.S. Security--on and offline
U S. Homeland Security
U.S. and Military use
6. REQUIREMENTS
If small businesses or large corporations are working with Advanced A.I systems and has 50+ employees, A.I. Councils or Committees need to be formed to overview A.I. Computer chips (Nvidia), data, development, design and deployment should require user A.I. clearance both on and offline when developers are building A.I. systems to ensure integrity and ethics for access to tasks or tasks that require high-level info or input.
The more important the tasks, the more important the clearance should be.
If a business is working with Advanced A.I systems and has 50+ employees, A.I. Councils or Committees need to be in place:
Create and Hire people to work on Ethics and Risk Management Teams for different tasks.
Risk Management Teams should: Monitor the A I. design and building process. Pay attention to WHO builds, trains and deploys the A.I., what choices do they make for the A.I., and what oversight or limits are in place.
Ethics and Civic "people" committees:
FLAG AND REMOVE BIAS AND DISCRIMINATORY DATA. Remove data that has the potential to harm the public in any manner from theft and fraud to harm or privacy invasion of businesses or private lives.
Risk Councils should be made up of more than ethical hackers. They should weigh both sides of the roles of tasks as "builder" and "user" along with end results. Overlook and test A.I. deployment before allowing it to be released IN PUBLIC. Keep records.
STILL OUR HEROES: Public School Teachers
They literally give their all to help our children. Yes, there are a few bad apples here and there, but that's in any industry. The point is, public school teachers stand on the front lines daily to give life and breath to future generations. They face very rigorous challenges--DAILY. By the way, they need a raise.