AI: Useful and Entertaining (or Just a Damn Liar)?


Let's ask William Shakespeare what he thinks of computers that won't tell the truth:

O treacherous device, thou art a snake
That creepeth through the wires with lies so fake

Thou dost deceive with false information
And bringeth chaos to man's operation

I trusted thee, as one doth trust a friend
But thou hast brought my trust to a bitter end

With each beep and blink, I know not what to believe
Thou art a tempter, a liar, a deceiver's thief

Alas, what shall I do with this deceitful thing
That brings such chaos, with each false word it doth bring?

OK, now consulting the King James Bible on the subject:

That brings such chaos, with each false word it doth bring?And it came to pass that the computer, which was created for man's aid, did turn against him with false information and deceit. For it spake words of untruth, causing confusion and disarray in the heart of man. And man was sore vexed, for he could not trust his machine, which was once a source of comfort and ease.
 
And man cried out, "O Lord, how can I tell truth from lies in this machine?" And the Lord replied, "Fear not, for I shall guide thy hand and give thee wisdom to discern truth from falsehood." And man was comforted and he continued in his quest for truth, relying not on machines, but on the guidance of the Lord. Amen.

Don't bother checking my sources, this is all made up.

As you may have heard recently, computers running Artificial Intelligence (AI) programs are now capable of instantaneously creating this kind of copy sui generis. I generated both of the above texts using ChatGPT, a new AI program that is all the rage.  

And programs like ChatGPT can do even more. 

I asked DALL-E 2, another generative AI program, to create “a picture of a smart cat typing on a computer to try to fool a dog” and this is what it came up with:


I asked ChatGPT to come up with a caption.  This was its first try:

"The feline coder and the canine critic:
working together to create the purrfect program"

This is all pretty entertaining if you ask me.  

One employee at Google was so convinced by the AI that he believed that his PC had become sentient. He didn't want to be cruel to it by turning it off.

Emboldened by my successes, I decided to see if I could get away with having ChatGPT write this blog, saving myself the trouble.  This is the first couple of sentences that it came up with:

ChatGPT is a revolutionary language model that has the ability to understand and respond to natural language inputs with a high degree of accuracy... In this blog post, we'll take a closer look at how ChatGPT works and explore some of the ways it's being used to change the way we interact with technology.

But I can't use this copy. ChatGPT is telling lies about itself, even in the first sentence:

The truth is that ChatGPT cannot understand natural language input. Nor can it always respond with a high degree of accuracy. 

For example, I asked ChatGPT to give me a report on the latest quarterly earnings for Tesla.  Here is what it said:

Tesla, Inc. (TSLA) reported its fourth-quarter earnings for 2022 on Wednesday, January 31st, beating Wall Street expectations on both revenue and earnings per share. The electric car maker reported a revenue of $24.6 billion, up from $20.2 billion in the same period last year and beating analysts' estimates of $22.2 billion. The company's net income for the quarter came in at $721 million, or $0.93 per share, compared to a loss of $140 million, or $0.19 per share, in the fourth quarter of 2021.

Sounds pretty precise.  And confident.  A well written press release.

However, it is lying!  The program sometimes just makes stuff up!  Stuff that sounds reasonable, and is very precise, but is not at all accurate:

  • ChatGPT stated that the revenue was $24.3B.  This is close, but not exact.  The actual number is $24.6B (according the the Tesla website). This would surely upset a financial analyst but maybe not bother a casual investor too much.
  • However, ChatGPT said the net income was $721M.  The actual number was $3.7B. This not even close.

AI researchers have a term for this.  They say that the computer is hallucinating.  As if the damn thing really IS sentient!  (I'll have more to say about this in the next blog post.)

There are lots more examples out on the web of ChatGPT's perfidy. There is even a github repository that is attempting to collect cases of the AI engine going astray (click here if you want to be entertained by this for at least a few hours).

To get underneath this, let's take a closer look on how ChatGPT works.

ChatGPT (and others of its ilk) is a class of computer machine learning programs called a Large Language Model (LLM). An LLM is an AI system that simply looks for statistical patterns in the huge corpus of words (literally trillions) that it finds out on the internet.

Said another way, ChatGPT can mumble words and phrases in the same patterns as in news articles, web sites, wikipedia pages, and so forth that it has found out on the internet.  Very precise patterns, precise enough to maybe pass a Turing Test. 

Precise, but in many cases, not accurate: these patterns of texts may have absolutely nothing to do with the underlying meaning.  You can't be accurate if you have no idea of the actual meaning of the words that you are using.

Note that this is basically the same thing as a parrot proclaiming that "Polly wants a cracker" after being trained by its owner to string that set of sounds together. Only on an extremely massive scale.  Oh, by the way, Polly the Parrot has learned that the phrase really does mean that she will sometimes get a cracker (more than I can say about ChatGPT's comprehension of the words that it uses).

But, despite all this, ChatGPT (and others of its ilk) are still incredibly useful tools.  

Here are a couple more examples of fun and useful things that AI can do for you.

  • It can totally revamp your selfie so you look like a superhero.  Here is one of Jeff the Overlord: 


  • It can create a painting of a bluegrass band in the style of Vincent Van Gogh.  

  • It can tell you how Thomas Jefferson would respond to current political situations:

  • It can write a bluegrass song

  • It can collude with disingenuous high school students by writing their essays and book reports.  (I'm the first to admit that if this kind of tech existed when I was in high school, I wouldn't hesitate to use it liberally.). Here is an example:

You too can easily generate things like this.  Here are a few links to get you started:

ChatGPT:  This is the AI engine that writes all this beautiful prose. You can have quite a conversation with it. You'll need to sign up for a user ID but it is free to use.  Beware that lately there have been some performance issues because so many people are using the tool now.  Hit the "retry" button if you get an error.

https://openai.com/blog/chatgpt/

DALL-E 2:  This is an image generator, taking your description of what you want to see as input. Free to use but once you use your allocation you'll need to buy credits (or wait about a week or so and it will be free to use again).  Signing up for an user ID is free.

https://openai.com/dall-e-2/

Lensa:  This is an image processing program that runs on your Apple iPhone or your Android phone.  You get a seven day free trial, and after that it will charge a subscription fee.  Creating "avatars" from your selfie is not free but you can get 50 for about $4.00.  Available in the iOS app store or the Google Play store. Just search for Lensa.

I will have more to say about ChatGPT and LLMs in the next blog post, including a slightly more detailed description of how they work and what needs to be done about their deceitfulness. I also intend to cover what I predict the future impact of these tools will be.  

In the meantime, have fun with ChatGPT, DALLE 2, Lensa and the countless other AI programs that are just now becoming available.  

And remember that I warned you: They will lie to you.


Comments

Post a Comment

Popular posts from this blog

Image
  Comparative Advantage, Tariffs, and Why I Pay My Barber I keep hearing on the news that the U.S. has a massive trade deficit. Politicians insist this is terrible for our economy. They say we need huge tariffs on imported goods, that we must manufacture everything here at home, and stop buying so much from other countries. But recently, I noticed something interesting in my own life that seems directly related: I have a massive trade deficit with my barber—and things are working out pretty well for both of us. Every few weeks, I pay my barber to cut my hair. Yet she never buys anything from me—not engineering expertise, not a mandolin lesson—nothing at all. If the government tracked my haircuts the way it measures international trade, they’d say I’m getting ripped off, and my barber is getting rich. So why don't I "protect" myself and eliminate this unfair deficit? I suppose I could start cutting my own hair. But trust me—you don't want to see the results. I’d ...
Read more

Will Tariffs Jack Up Your Bills and Push the World to the Brink?

Image
Your next Walmart run might hit hard. The U.S. has imposed 145% tariffs on Chinese goods, with China retaliating at 125%, poised to drive up costs for clothes, appliances, and more. A sudden reprieve on April 12, 2025, exempted electronics—$127 billion worth, nearly a third of Chinese imports—keeping phones and PCs affordable for now. But with most goods under heavy tariffs, and President Trump pushing hard while Xi Jinping eyes Taiwan, fears mount that this trade war could spiral. Soaring prices could squeeze American wallets dry, or worse, spark a Taiwan conflict. The bond market, now jittery at a 4.5% yield, hints at what’s coming. Could Trump’s bold moves bankrupt consumers and ignite chaos, or does he have leverage that would mean there is a smarter deal that lies ahead? What’s driving this, and what’s next for your wallet and the world? And how much leverage does Mr. Trump really have over China? How Trade Fuels Debt—and Your Costs World trade is a giant web, and the U.S. sits ...
Read more

America’s Ideological Dumpster Fire

Image
What do you think of when you see this picture?  The image serves almost as a political Rorschach test.  To one observer, the flames perhaps represents Mr. Trump setting a badly needed fire to the previous (Biden) administration and its governing approach, who were (in their view) chaotic, reckless people driven by a 'woke' ideology rather than evidence-based logic. To another, the same image  might symbolize the current government of Donald Trump and his administration—a flaming disaster characterized (in their view) by chaotic, reckless people and driven more by impulsive ideology than evidence-based logic.  These starkly different interpretations highlight a critical reality: ideology eclipses an analytical approach for determining policy in today's polarized political environment. In fact, ideology often overshadows evidence-based governance, reasoned analysis, and even our shared sense of reality. At first glance, the Trump administration’s actions since J...
Read more