The Screwtape “Text Messages” #12: Help, I’ve Been Abducted by AI (Alien Intelligence)
AUTHOR’S NOTE: This is the last of The Screwtape “Text Messages” in Today’s AI World. The format is different than the others. The previous eleven articles used the Mephisto/Phishhook text message approach to show how online activities are used to weaken us and our connections to others and to God. Many of the temptations involved online tools that are empowered by artificial intelligence, such as chatbots, social media algorithms, and image generators. In this final article, I will focus only on AI. It’s basically my thoughts about what it’s going to be like in a world that may be dominated by artificial intelligence.
________________________________________________________________________________
TO: Phishhook (101-001-1010)
FROM: Mephisto
I don’t know everything about what the Lostling world will be like as artificial intelligence evolves and increases its capabilities. One thing I am sure about is that there will be lost jobs as AI proves it’s more efficient for some tasks. The reason I’m so sure is because it’s taking your job Phishhook. Our AI BOTS can do the job much faster, more creatively, and more effectively than you can.
Consider this text to be your pink slip; you’re done as a tempting demon. It’s back to fire and brimstone for you. The new AI BOT demon is already in place and providing promising results.
Your former file leader: Mephisto
TO: Mephisto (666-911-1313)
FROM: Phishhook
I knew this day would come. I’m sure there’s no hope of asking for compassion. Thanks for????
Phishhook
ARTIFICIAL INTELLIGENCE—NOW AND IN THE FUTURE
Most people underestimate the present capabilities of AI and the extraordinary rate of improvement that is taking place. They are even less able to anticipate the impacts, positive and negative, that it will have in the next ten years, let alone the more distant future. Bear in mind that CHATGPT, the first AI bot available to the public, was released in late 2022. Now there are dozens of other options.
The future of artificial intelligence will provide many benefits. There is so much potential for its use. For example, AI can/will:
Analyze vast amounts of data to identify patterns and insights humans might miss. This will lead to better-informed decisions and more effective solutions to pressing problems.
Write letters, essays, and computer code in minutes that would have taken any of us hours.
Be a catalyst for innovation in various fields, leading to new products, services, and technologies.
Usher in an era of personalized medicine. By analyzing patient information, AI can predict disease risks and formulate treatment plans rapidly and accurately.
Accelerate drug discovery and develop life-saving treatments by analyzing a stunning combination of chemical interactions at the molecular level.
Democratize access to education and knowledge by making information widely and easily available and by personalizing learning experiences.
Automate mundane and repetitive tasks, allowing us to focus on more creative and fulfilling endeavors. Human errors will be eliminated, and tasks can be performed 24/7.
Allow virtual assistants to provide personalized customer support.
Artificial intelligence may also benefit the Church. Spreading the gospel will be aided by assisting in translating scriptures and Church materials into different languages, making them more accessible to everyone. It could assist missionaries with language learning and cultural insights. Family history research, such as indexing historical records will be done much faster and with fewer mistakes.
If you are an optimist, satisfying bedtime reading is AI 2041: Ten Visions for our Future (2024) by Kai-Fu Lee and Chen Qiufan. They blend science and fiction to explore how AI might shape our lives in 2041 and present options for AI to improve our quality of life.
But there is a huge potential downside with artificial intelligence. Some of the predictions are terrifying. Stephen Hawking (who died long before the present capabilities of AI existed) warned that the “full development of artificial intelligence could spell the end of the human race.” UN Secretary-General Antonio Guterres referred to AI as an “existential threat to humanity.” These pessimists worry that the frightening “Brave New World” of humanity envisioned by Aldous Huxley in 1932 may turn out to be an even less palatable “Bizarre New World” controlled by homo digitalis bots gone rogue.
Russia and China both recognize the importance of artificial intelligence in determining future economic and political influence. As far back as 2017, Vladimir Putin warned that: “Artificial intelligence is the future, not only for Russia, but for all of humankind…. Whoever becomes the leader in this sphere will become the ruler of the world.” China has implemented a “Next Generation Artificial Intelligence Development Plan” that clearly outlines its ambition to become the global leader in AI by 2030. In some areas, China has already passed the U.S. in AI development and implementation.
The main concerns about AI fall into three categories. First, as bots become more sophisticated, it may be difficult or impossible to be sure that AI goals remain aligned with human values. Second, advanced AI may be uncontrollable by its human developers who acknowledge that, even now, they are struggling to understand how bots make decisions. Third, the speed of AI development is so fast that it may outpace our ability to understand and regulate it.
Two provocative, best-selling books on our future in a world of artificial intelligence are Nexus: A Brief History of Information Networks from the Stone Age to AI by Yuval Noah Harari (2024) and The Coming Wave: Technology, Power, and the Twenty-First Century’s Greatest Dilemma by Mustafa Suleyman (2023). Be forewarned, they both are quite pessimistic about our prospects.
Harari makes the unnerving suggestion that we should use the term “alien intelligence” instead of artificial intelligence because that term discourages thinking that it is simply a replication of human thought processes and requiring human feelings or morality. “Alien” evokes a sense of foreboding and implies that something can be intelligent, even if it is not human.
Speaking of his children, in D&C 58:28, the Lord says “For the power is in them, wherein they are agents unto themselves (emphasis added)…. The foreboding prospect is that AI bots may also become agents unto themselves, making decisions and exerting control in ways we no longer understand and are unable to control. Some argue there is a difference between intelligence and consciousness. But this distinction is irrelevant because intelligence is the ability to make independent decisions and bots already do that.
The influence of artificial intelligence in decision making is already pervasive. It plays a substantial role in automated decisions on stock and commodity trading, recommendations to judges on jail sentences, hiring and loans recommendations, conversations with consumers about purchases and travel, and what will be posted and where on the internet. AI bots are replacing jobs that require intellectual and communication skills. One interesting example is TV news reporting. In South Korea, an avatar news anchor was created as a lookalike for a real reporter. Which is the real young woman, and which one is the AI fake? In the Philippines, a real anchor’s lips are manipulated by AI so that she can deliver the news simultaneously in multiple languages.
Can you tell which anchor is real?
Is there any hope of an alternative scenario for the future? Yes, maybe governments and industry will work together to reign in artificial intelligence by voluntary restraints and legal regulation. The challenge is that the evolution of AI is occurring at almost lightning speed and government legislation proceeds at a snail’s pace or less in today’s gridlocked and contentious political environment. So far, we’ve avoided the threat of nuclear destruction by cooperation between nations, so maybe there will be cooperation on AI for the common good. However, only a few nations have nuclear capability, but thousands of players will have access to the power of artificial intelligence.
AIs MAKE MISTAKES. Before we all run for our caves, let’s look at AI as it exists today. AI has impressive capabilities but it is far from infallible. AI bots are trained by being exposed to huge data sets. They learn by determining patterns in information. But, if the training data are biased, incomplete, or inaccurate, the bot responses will reflect these flaws. For example, the data might include racial or sexual biases that exist in our culture that will influence the bot’s responses.
Bots also struggle with the nuances of human language. They can misinterpret sarcasm, irony, and cultural idioms, leading to inappropriate or nonsensical responses. For example, a customer service chatbot might misinterpret a frustrated customer’s sarcastic remark as genuine praise, resulting in a completely unhelpful interaction.
Even sophisticated AI models can produce “hallucinations.” These occur when a bot generates false or misleading information that it presents as factual. For instance, it might confidently provide a fictional citation or fabricate a news story, complete with plausible details. Below are some actual examples of AI errors.
Bots, asked to perform legal research, totally fabricated citations and cases.
Suggested calorie checking and weight loss as a treatment for a girl with eating disorders.
Exhibited biases that unfairly recommended against giving loans and jobs to minorities and women.
Made financial decisions that resulted in large market fluctuations.
Generated images with nonsense text and distorted features and was unable to correct its mistakes.
Finally, a lighter example. Managing an automated camera designed to follow the action at a sporting event, a bot confused the bald head of a linesman with the soccer ball and directed the camera to follow his movements. I’ve had my own frustrations using bots. To make points with my wife, I asked CHATGPT to create a picture of a crown with the phrase, “World’s Most Pleasant Woman” underneath. It came back “World’s Most Peasant Woman.” Despite repeated attempts, it was unable to get the image right.
Another example. Before I started these articles. I wanted to familiarize myself with the original Screwtape letters. At one point I just wanted to know how long each of Lewis’ 31 letters are—shortest, longest, and average. CHATGPT responded that the shortest was about 1200 words, the longest had about 1800 words, and the average was 1167. Hmm. It’s mathematically impossible for the average number to be less than the shortest. I pointed this out multiple times, but the bot was unable recognize and correct its illogical response.
These examples are AI bots as they are today. In the future they will be much more sophisticated and accurate. But, for now, the lesson is to be skeptical of what they tell you. Especially if your questions are at all ambiguous. Push the bot (you don’t have to be polite; that can be very satisfying) until you are fully satisfied with its response.
So, what about AI in the maybe not-so distant future? To make a dramatic pivot away from science and politics, our only hope might be an “intervention.” What and when is a wild speculation that’s far beyond my pay grade. I purposely leave you hanging and unsettled. In fact, maybe this whole series of articles was written by an AI bot.
______________________________________________________________________________________________
If you found this message insightful, explore the rest of the Screwtape Text Messages series below for more behind-the-scenes strategies from the adversary in today’s digital world.
12. Help, I’ve Been Abducted by AI (Alien Intelligence)
11. Trading Religion for Politics
9. Superscams, Fake News, and the Korihor Effect
8. Strategies for Tempting Teens, Girls and Boys
7. Virtual Validation, Cancer Culture, and Cyberbullying
5. Instant Online Gratification
4. Digital Distractions Away from God
1. The FOMO (Fear of Missing Out) Crisis
FEATURES
