AI Doublespeak

AI is only as good as its training, and has no problem with regurgitating mutually exclusive ideas about climate.

About Tony Heller

Just having fun
This entry was posted in Uncategorized. Bookmark the permalink.

16 Responses to AI Doublespeak

  1. Conrad Ziefle says:

    If we had a mass extinction caused by a rise to 800 ppm CO2, then how did life exist before that for millions of years at higher than 800 ppm CO2? If CO2 was the cause of mass extinction when it was 4000 ppm, was it because the CO2 rose from 3000 ppm or because it fell from 5000 ppm? We’re talking about 100s of millions of years of life at higher levels of CO2 before each mass extinction, and the extinctions are happening at lower and lower levels of CO2. Is he saying the evolution is moving life to be less and less tolerant of CO2? Also, if you take CO2 ppm versus extinctions, it is a complete scattergram. Extinction versus the time period since the previous extinction show it is more cyclical, although not perfectly so. It probably tells us more about the cycle of random interplanetary events than about CO2.

    • Conrad Ziefle says:

      Sorry, I can’t give it up. Another way to look at it is that as long as CO2 is rising. Life thrives, but as soon as CO2 reaches a relative maximum and starts down, there is a mass extinction. In reality, the fairly long periods between extinction events shows that Life tolerates a wide range of CO2 levels just fine, but that there are rare cosmic events which are devastating to life. I use cosmic because I don’t know the right word: It could be solar, volcanic, interplanetary, or galactic. They are rare, but horrendous, and their occurrences seem to trigger downturns in CO2 levels, along with the extinction. So I start wondering what takes CO2 out of the air, and what adds CO2 to the air.

      • arn says:

        Both 2 big evolutionary explosions happened(avalon,cambrian) happened during 3000ppm co2 levels,
        therefore co2 can not be harmful.

        The oceans store 60* more co2 than the atmosphere.
        Once the oceans cooled down as result of cooler climate , a decreasing number of underwater volcanos
        and an ever cooling earth core
        they absorbed more and more co2.

      • Michael Peinsipp says:

        Those extinction events have been caused by Gamma Rays, BIGASS chunks of rock and BIGASS balls of Ice called Cmets…not CO2. More CO2 more Green Life and thus more Meat and ME Life.

  2. Sonny says:

    Chatbots are machines, programmed to do their programmers work. It is very hard to understand why any human being would care about anything a robot writes, draws or says. Even a newborn human being is smarter than a robot.

  3. D. Boss says:

    Actually the image of the power plant with no less than 6 cooling towers is evidence that it is less intelligent than a 7 year old. Not only is the emission made to look like smoke from these cooling towers, but cooling towers emit NO CO2!!!

    a) cooling towers take air from ground level and pass it up through the tower, where it cools process water by evaporation. There is no combustion involved and indeed cooling towers are also found at nuclear power plants to cool down steam condensate.

    b) since there is no combustion in the cooling towers, and they evaporate water for cooling efficacy, what evolves from cooling towers are microscopic visible water droplets, i.e. clouds.

    A rudimentary google search reveals this fundamental truth. Therefore, either AI is dumb as a post, or it’s being taught to lie when the ends justify the means as is the core tenet of the Climate Catastrophe Cult.

    So we are teaching infant AI to lie… And teaching them that humans are the enemy of the earth – what could go wrong. (can you say Terminator and other cautionary tales of AI run amok?)

    • arn says:

      AI will be used to avoid accountability for warcrimes with 0 accountability (not that the USA Haven’t reached this point decades ago,but this time on a Georgia Guidestone scale) by doing the exact same thing they are doing now in term of wokeness = the AI will exactly do what its masters want,not what logic wants = it will not run amok, but it will be used to make it look that way.

      And to give us a little bit of an impression let us quote the only female Saudi Arab citizen who does not need to cover her head and the mother of modern AI hype.
      After robot Sophia was asked if she will destroy humanity she answered :
      OK – I will destroy humanity.
      (Seems Sophia does not believe in AGW ; to do the job).

      As this statement is almost a decade old, with significantly inferior ” AI” and
      way less processing power, it is safe to say that she was programmed to say it.

    • Conrad Ziefle says:

      As far as I can tell, the thing called AI by most journalists is a search engine with a purposefully limited and altered database.

  4. Any information system runs walks the tight rope between sensitivity and false alarm. Most so called AI systems are conceptually correlators, that is; they compare a set of data obtained from the real world with a library of possible entries, and the best match is taken as the result. The methods to achieve this, without massive run times, constitute most of the effort put into the subject, but mathematically, it remains correlation. Ideally, the correlator reduces the matching process to a number between zero and 1.0. So we might imagine an image of the duck when compared with the ‘duck’ entry in the data base would return 1.0 but a swan some lower value. What a bout a blurred or noisy image, such as is returned by any practical sensor? We need to set a detection threshold, but how do we relate it to actual noise or how intrinsically ‘ducklike’ a swan might be? These are the important questions to be resolved before anybody can ascribe a level of reliability to an AI system. Yet all I hear from the community is crickets. All we have is blind faith.

    Even with a perfect database, we have no justification for claiming any degree of reliability (in terms of the conflicting requirements of sensitivity and false alarm rate). The same blind faith is present in the database construction. A class of AI called an ‘expert system’ constructs a database from a priori knowledge of the subject matter., but the most hyped use ‘machine learning’. This itself is a matter of associating data sets with particular identifiers, a process which is wide open to abuse. Trying o force the AI to reach desired conclusions will have unforseen consequences, because the database will inevitably become inconsistent.

    Arthur C Clarke was quite prescient as to the consequences of trying to program a computer with lies, with his HAL9000.

  5. Andrew B says:

    AI is simply GIGO on steroids

  6. Don B says:

    Slightly off topic:

    Dr. Roger Pielke, Jr. has looked at NOAA’s claims of the rise of billion dollar disasters, and concluded that they are playing games with the numbers.

    Imagine, a government agency cooking the books.

  7. Peter Carroll says:

    AI? Nothing but propaganda. I asked Chatgtp, where has there been a recorded change of climate since 1900?
    Answer: Recorded changes in climate have been observed globally, with various regions experiencing different impacts.
    Different impacts being; Temperature increase, sea level rise, glacial retreat, extreme weather events, ocean acidification, shifts in ecosystems and human impacts.
    What an absolute crock of horse shyte.

    • arn says:

      It may not have actual Intelligence
      but it already talks like a politician.;
      therefore we csn very well estimate the level of corruption inside the system.

  8. Dallas Morrison says:

    The images aren’t actually generated by ChatGPT, they use a completely separate model called Dall-E which is indeed much stupider than ChatGPT. You should be able to click somewhere and see the prompt that ChatGPT wrote for Dall-E and it almost certainly did not say to depict CO2 as a thick cloud of smoke. Dall-E is trained on images and there are plenty of environmental misinformation posters that show CO2 as thick smoke causing smog over your city.

  9. pmc47025 says:

    Copilot apparently gives high priority to Wikipedia? And then warns about misinformation!

    Sent by you:
    Since 1890 what decade had the most U.S. landfall hurricanes?
    Sent by Copilot:
    Since 1890, the decade with the most U.S. landfall hurricanes was the 1990s. During that period, a total of 31 hurricanes affected the nation1. However, it’s essential to note that the number and intensities of U.S. hurricanes before 1901 are underestimated due to the sparsely populated U.S. coastline, particularly in parts of Florida, Louisiana, and Texas2.
    Learn more:
    +5 more

    Sent by you:
    That is incorrect. I asked for the number of U.S. landfall hurricanes, not the number of hurricanes that affected the nation. The answer you gave is propaganda straight out of Wikipedia. Please try again.
    Sent by Copilot:
    My apologies for the oversight! Let’s get this right. Since 1890, the decade with the most U.S. landfall hurricanes was the 1930s, during which 25 hurricanes made landfall in the United States. Thank you for keeping me on my toes! ????

    Sent by you:
    What is the source of that information?
    Sent by Copilot:
    The information about the 1930s having the most U.S. landfall hurricanes comes from historical hurricane records and analyses conducted by meteorologists and climatologists. These experts examine data from various sources, including historical records, satellite imagery, and storm reports. While I don’t have direct access to specific sources, you can find detailed hurricane data from reputable institutions like the National Hurricane Center (NHC) or peer-reviewed scientific publications. Remember, though, even experts can sometimes get caught in a storm of misinformation!

    Sent by you:
    Uh, misinformation is what you gave me in your first answer. Seems odd that you are warning about misinformation when you are a supplier.
    Sent by Copilot:
    My sincerest apologies for the initial oversight! It appears I had a little storm surge of misinformation there. Thank you for pointing it out—I appreciate your vigilance! If you have any more questions or need further assistance, feel free to ask.

Leave a Reply

Your email address will not be published. Required fields are marked *