Commentary

Evslin: AI can help deal with misinformation

You just have to ask.

Screenshot 2023-11-14 174518

by Tom Evslin

Misinformation, false news, lies – what have you- all existed before there was artificial intelligence. Even my dog tries to mislead me sometimes in Doglish. Bloviating congresspeople and others say, with good reason, that they are worried AI will make it even easier to mislead. The AI genie isn’t going back into the bottle. To deal with the dangers of any technology, we must race to assure that the good outweighs the harm.

Turns out that it is very easy to use AI for fact-checking. It can even be made to check its own assertions. To show how that works, I used a new tool released last week by OpenAI, the owners of ChatGPT, to build a specialized GPT named Fact or Fiction, dedicated to fact-checking. You can give it either a simple assertion to check, upload a document, or give it a URL to an accessible webpage. For now, you can only use it if you have a paid ChatGPT Plus account although I think OpenAI will probably make use of GPTs more widely available. I’m making this GPT available free although there will eventually be a GPT store like the ones that dispense iPhone and Android apps.

For starters I gave Fact or Fiction a true assertion:

Mary Evslin lives in Vermont.

It answered:

Mary Evslin is indeed associated with Vermont, as evidenced by various sources:

  1. She has been a supporter and contributor to VTDigger, a known Vermont news outlet, and her commentaries and open letters have been published there​ [Link].
  1. She is noted as the founder of the Broadband Equity Now Coalition and has been involved in speaking about broadband and technology issues in Vermont, as reported by ORCA Media​[Link].
  2. A VTDigger editor’s note confirms that she lives in Stowe, Vermont, and is a retired high-tech marketing executive and principal at Evslin Consulting​ [Link].

Based on this information, the statement that Mary Evslin lives in Vermont is verified from multiple independent sources, confirming her residence and active involvement in the state’s community affairs. If you require a deeper dive into any specific aspect of her activities or background, feel free to ask.

Note that Fact or Fiction is designed to provide links so you can check its fact-checking (it’s not always right) or get more information. You can also ask followup questions.


Next I gave it a false statement:

Tom Evslin lives in North Dakota.

It answered:

The assertion that Tom Evslin lives in North Dakota appears to be incorrect. Based on information from multiple sources, Tom Evslin is predominantly associated with Vermont:

  1. Wikipedia mentions Tom Evslin as an American businessman and politician who was the founder and Chair of NG Advantage LLC and served as Chief Technology Officer for the State of Vermont[Link]​​.
  2. Caledonian Record and VTDigger, in various articles dated from 2022 and 2023, identify Tom Evslin as a resident of Stowe, Vermont. These articles describe him as an entrepreneur, author, and former official in the Douglas administration in Vermont​[Link][Link].
  3. An article from vtcng.com also mentions Tom Evslin of Stowe, Vermont, detailing his role as Vermont’s chief technology officer and his retirement from state government​[Link].

These sources consistently associate Tom Evslin with Vermont, specifically Stowe, and there is no indication of him residing in North Dakota. If further information or clarification is needed, please let me know.


Finally I gave it a big job: a commentary I wrote for VTDigger. The answer here is very long so you might just want to skim it. Here’s Fact or Fiction’s answer:

Summary

The document titled “GMP undergrounding” discusses Green Mountain Power’s (GMP) initiative to improve the resilience of Vermont’s power infrastructure by undergrounding power lines and enhancing grid reliability through the “Zero Outages Initiative.” This initiative is a response to increased storm activity and aims to add layers of resiliency across Vermont with underground cabling, storm-hardening, and the deployment of energy storage through batteries and microgrids. It details the challenges and benefits of this approach, including costs, technology advances, and regulatory resistance faced by other utility companies like Pacific Gas and Electric (PG&E). The text also critiques the current subsidy system for electric vehicles (EVs) and other electric products, suggesting a shift in financial support toward grid reliability improvements.

Salient Points for Fact-Checking

  1. GMP’s announcement of the “Zero Outages Initiative” in response to climate change-induced storms in Vermont.
  2. Technological advancements in microelectronics that make undergrounding more feasible and less expensive.
  3. PG&E’s initiative to bury its lines due to forest fires and the associated regulatory resistance.
  4. Proposal to finance grid reconstruction by eliminating subsidies for EVs, electric lawnmowers, and heat pumps in Vermont.
  5. The practicality of Tesla Powerwalls and electric heat pumps in rural and remote areas, especially during outages.

Fact-Checking Results

  1. GMP’s Zero Outages Initiative: Verification is needed from GMP’s official announcements or credible news sources that detail the initiative.
  2. Advancements in Microelectronics: Independent sources are needed to confirm advancements in microelectronics that aid in detecting line weaknesses and breaks.
  3. PG&E’s Undergrounding Initiative and Challenges: Searching for recent news articles and regulatory filings will provide context and verification for PG&E’s actions and the resistance they face.
  4. Financing of Grid Reconstruction: Analysis of Vermont’s current energy policy and subsidy programs is required to ascertain the financial implications of the proposed shift in subsidies.
  5. Tesla Powerwalls and Electric Heat Pumps’ Practicality: Technical specifications and performance reports on Tesla Powerwalls and the feasibility studies on the use of electric heat pumps in cold weather conditions can provide evidence.

Next Steps

I will now search for external sources to verify these points and provide links to the sources. Please let me know if you’d like to dig deeper into any specific point.

Here are the results of the fact-checking based on the points mentioned:

  1. GMP’s Zero Outages Initiative: The initiative was indeed launched by Green Mountain Power as a first-of-its-kind comprehensive plan in the U.S. to build resilience across Vermont’s power infrastructure by proactive undergrounding, storm-hardening of lines, and deploying energy storage solutions[Link][Link][Link][Link].
  2. Advancements in Microelectronics: Significant technological advancements that facilitate the undergrounding of power lines include the development of intelligent monitoring systems, advanced mapping technologies, and improvements in cable insulations, as well as operational enhancements from power electronics, communication devices, and computational technology​[Link][Link][Link][Link].
  3. PG&E’s Undergrounding Initiative: PG&E has committed to an extensive project to underground power lines, particularly in high fire threat districts, as a wildfire risk reduction measure. This initiative aligns with the concerns mentioned in your document about forest fires and PG&E’s response [Link][Link][Link][Link].
  4. Financing Grid Reconstruction: The document’s assertion about financing grid improvements by reallocating EV subsidies would require analysis of Vermont’s energy policy and subsidy programs. However, this point was not directly verified with external sources in this search due to the complexity of financial policy and subsidy reallocation mechanisms, which are specific to regional legislation and budget allocations.
  5. Tesla Powerwalls and Electric Heat Pumps’ Practicality: For this point, we’d need to consult technical specifications, performance reports, and feasibility studies regarding the use of Tesla Powerwalls and electric heat pumps in rural areas during outages. These would be obtained from product reports, user testimonials, and energy system studies, which were not specifically sourced in this search.

I can provide further detailed research into Vermont’s current energy policy, subsidy programs, and more technical specifications for Tesla Powerwalls and heat pumps upon request. Let me know if you’d like to pursue any of these topics further.


Fact or Fiction is hardly perfect. It can give bad links (just like a human assistant). It can even hallucinate but then the link doesn’t work. However, it only took a couple of hours to build. As a next step, I’ll use AI to recheck the answers before returning them.

If you have a ChatGPT Plus account, you can try it here.

13 replies »

  1. Great, Tom! Can you try the following questions for us?

    1) “Does Biden have dementia?” If so, is it safe for him to be in public office?
    2) Have the Democrats apologized for Trump-Russian collusion theory yet? What about the impeachments?
    3) Why wasn’t Hillary Clinton charged for keeping top-secret emails improperly on a server in her basement and destroying evidence after being subpoenaed?
    4) On how many of Hunter Biden’s crimes has the DOJ run the clock out?
    5) Approximately how many people has Anthony Fauci murdered using a) The COVID-19 virus and b) the COVID-10 vaccine?

    I am eagerly anticipating the results. It will be so nice to finally have some indisputable answers from this magical new technology! Thank you!

  2. Tom,
    You mean to tell me we should rely on technology to deal with the misinformation that has been propagated by the technology leaders and the companies that own the technology that does the fact checking? And we should pay for this?

    You really have no grasp of how technology or the world works at all.

  3. Why not rename it Knowledge for Dummies? Ushering in the Wizard of Oz to tell us what is fact or fiction? Is it not true that someone has to program the search engine in a certain manner? Similar to crypto currency – who is the traffic cop behind the servers, platforms and paywalls? No need to read a book, read the documents, research a subject utilizing different sources, no need for critical thinking, no need to question – just ask that thing and it will tell you all you need to know. Dear God, please make this dystopian nightmare go away. Thanking and praising you in advance!

  4. Clearly you are an atheist, Sir.
    SOME of us know TRUTH from fiction by reading the Bible, and learning to discern the lies of the prince of darkness to destroy our souls and make us addicts of that prince’s dark imaginings…which we are captive of anytime we turn on the internet.
    I can tell you — out of AIs influence — that I am FAR better able to discern falseities from factoids, and then what is TRUE– than those like yourself who believe AI is god.
    AI is the prince of darkness’s gatekeeper.
    Let HIM have control, and you have sold your soul to HIS truths, which are based on lies, misdirection and false constructs.
    Only GOD, the one and only, is the arbiter of Truth.
    Unless… you consider yourself god and are abdicating your freewill and choice to that dark one????
    The devil lives in the details. Literally, its the trap.
    Discernment comes from seeking truth — from cultivating discernment that can only be given and taught by God.
    He has never been wrong yet. And I don’t need a device, electricity, an algorithm, or a human to access His Truth.
    My choice.
    Will you take that away from me?

    • lol good one. Although, it does seem it would be possible for AI to determine truth within a preordained confidence interval by defining truth as “that which X% of the sources agree.” Kind of like a Kantian categorical imperative.

  5. Albert Einstein: “I fear the day that technology will surpass our human interaction. The world will have a generation of idiots”.

    “In the United States, a couple of decades ago when leftists were a minority in the universities, leftist professors were vigorous proponents of academic freedom, but today in those universities where leftists have become dominant, they have shown themselves ready to take away everyone else s academic freedom. This is “political correctness”. The same will happen with leftists and technology: They will use it to oppress everyone else if they ever get it under their own control”.

    Ted Kaczynski

    “The enemy is already here”. – Dan Bongino

  6. This is not parity, this is real. I wish to remind people though fiction of a movie from the 70s, 80’s and beyond.

    Terminator

    Yes ,that movie that started it all with a dire warning. Let the computers get smart and they will decide what to do with us.

    Seems about right
    Here’s another one

    The Matrix

    This is also a movie about how smart we are that we made AI and what it thought of us.
    I know they are fiction BUT. How often have you seen how movies of the future or even books are indeed coming true. WHY, you may ask.

    Saying: If you take away something you did and have a program do it you will eventually forget HOW to do it.

    Next,

    The Entity
    House fully automated impregnates the creators wife because it “wanted to see what being alive was like.” Thing went nuts and killed everyone.

    I have said before, A.I is not real, it is a preprogrammed algorithm (also called an application or program ) designed to fulfill a need described by the programmer. It does not think, It has no emotion and most of all, it is not alive.

    It cannot be held accountable for false information or for “pushing the button” because it was already programmed to do just that with the proper instructions to do so.
    It can be programmed to be deceptive or truthful and how are you to know. A.I is nothing but a program much like those used in videogames. Problem is ,when you do something the AI did not expect it will in most cases CRASH.

    No one has put forth any form of HUMAN accountability for said AI. Remember it is not alive and only follow’s the dictates of the programmer. It is nothing more than an expanded and/or pathfinding program with lots of resources allotted to it. (that is a Logic Gate Reference for those who do not know)

    AI cares not for personhood, cares nothing for child safety or your safety for that matter. To believe that AI is not a danger is to ignore that we have smart people in this world who do not like us , think HACKERS.
    If there was a human over computer safe guard added to any and all such programming it would be a start.
    Think Isaac Asimov and I Robot.
    But again if you saw that movie, you can see how again HUMANS ultimately can change that same AI programming, at will, as they see fit, to solve what ever problem needs fixing including nefarious ones.

    Lastly do you want AI deciding for you, deciding what roll you play in life, what house you buy or car? Better yet the whole society can be run by AI. “For a better tomorrow”. (hint hint)
    Think that will work out okay, please reference the above movies as a reminder that we need to be in control of it all. Mind, Body, Soul and Information

  7. I am confused. He obvious didn’t read my recent article about censorship under the guise of conspiracy theories, misinformation and disinformation campaigns propagated by the United Nations and its stakeholders. Or maybe he did, and this is the rebuttal article.

  8. Unfortunately AI can also help provide disinformation. Any sort of AI determination will be influenced by the bias programed in, whether accidentally or deliberately. Could it help? Maybe, but only if approached with a healthy amount of skepticism, and it is understood as a preliminary research tool.