AI: Friend or Enemy
Helping other understand the impact of AI, I discuss history, current market state as well as encourage use of AI in 9/11/25 AI presentation MSU Lifelong Learning Center Fall 2025 curriculum.
9/11/2025 presentation in a 90 minute powerpoint session on Artificial Intelligence, AI, as one of the MSU Lifelong Learning Center’s Fall 2025 sessions of varying speakers and topics.
First the end; I’ve attached the slides as one file here
And the slides with the speakers notes as a second file here
Slides and comments as presented
Authors Note: These detailed notes take time and I’ll update in coming days till completed with individual links for YouTube video’s and other supporting files.
Following are the individual slides and some brief comments as presented.
Slide 1: ELIZA the first AI?
Joseph Weizenbaum, designed the ELIZA program to demonstrate the superficiality of human-machine communication. Weizenbaum spent the rest of his career warning against the overestimation of AI.
Slide 1: One early precourse today’s ChatBots was Eliza
And if you want to explore Eliza try this updated chat that shares the experience of Dr. Weizenbaum’s 1960’s program
"I HAD NOT REALIZED ... THAT EXTREMELY SHORT EXPOSURES TO A RELATIVELY SIMPLE COMPUTER PROGRAM COULD INDUCE POWERFUL DELUSIONAL THINKING IN QUITE NORMAL PEOPLE."
JOSEPH WEIZENBAUM, 1968
AI generated topic summary follows
Joseph Weizenbaum's ELIZA program is a well-known example of Rogerian psychology's influence on early artificial intelligence (AI). In 1966, Weizenbaum designed ELIZA to parody a Rogerian psychotherapist, surprising him and others by how deeply users emotionally engaged with the simple chatbot. [1, 2]
Rogerian psychotherapy
Developed by psychologist Carl Rogers in the 1940s, Rogerian therapy (also called person-centered or client-centered therapy) is based on the idea that people can achieve psychological growth if a therapist provides a supportive environment. Key components of this non-directive approach include:
Empathetic understanding: The therapist listens attentively to reflect and rephrase the client's thoughts and feelings back to them, which helps clients gain self-awareness.
Unconditional positive regard: The therapist accepts and supports the client without judgment, helping them feel valued and secure.
Congruence (genuineness): The therapist is open and honest to foster a healthy, open relationship. [6, 7, 8, 9, 10]
ELIZA's parody of a Rogerian therapist
Weizenbaum modeled ELIZA on the Rogerian method because its non-directive, reflective style was relatively easy for a computer program to imitate. Using a "DOCTOR" script, ELIZA would scan for keywords in a user's input and apply simple pattern-matching and substitution rules to turn the user's statements into a question. [2, 11, 12, 13, 14]
For example, if a user typed "My mother takes care of me," the program might identify "mother" and "me" and respond with "Tell me more about your family". If no keywords were found, ELIZA would use a generic, non-committal response. [15, 16]
Weizenbaum's critique of AI
Despite ELIZA's simple design, users readily projected human understanding and empathy onto the program, a phenomenon now known as the "ELIZA effect". This reaction deeply concerned Weizenbaum, who had designed the program to demonstrate the superficiality of human-machine communication. [2, 17, 18, 19, 20]
Weizenbaum spent the rest of his career warning against the overestimation of AI. He argued that the human capacity for empathy and genuine understanding could not be replicated by machines and was appalled at the suggestion that a simplistic program could stand in for a human therapist. In essence, the success of his parody revealed a profound human tendency to anthropomorphize and seek connection, even with a shallow technological imitation. [1, 2, 21, 22, 23]
AI responses may include mistakes.
[1] http://www.trs-80.org/talking-eliza/
[2]
[3] https://www.ncbi.nlm.nih.gov/sites/books/NBK589708/
[4] https://elliottcounselinggroup.com/what-is-person-centered-therapy/
[5] https://en.wikipedia.org/wiki/Person-centered_therapy
[6] https://link.springer.com/chapter/10.1007/978-3-030-57310-2_6
[9] https://imperfect.co.in/humanistic-psychotherapy/
[10] https://www.ncbi.nlm.nih.gov/sites/books/NBK589708/
[11] https://en.wikipedia.org/wiki/ELIZA
[12] https://web.njit.edu/~ronkowit/eliza.html
[13] https://arxiv.org/html/2406.17650v1
[14] https://www.psychologytoday.com/us/blog/our-new-discontents/202403/dont-be-fooled-by-ai
[15] https://cse.buffalo.edu/~rapaport/572/S02/weizenbaum.eliza.1966.pdf
[16] https://web.stanford.edu/class/cs124/p36-weizenabaum.pdf
[17] https://en.wikipedia.org/wiki/ELIZA_effect
[18] https://news.mit.edu/2008/obit-weizenbaum-0310
[20] https://www.sciencedirect.com/science/article/abs/pii/S0747563201000048
[21] https://spectrum.ieee.org/why-people-demanded-privacy-to-confide-in-the-worlds-first-chatbot
[22] https://ojs.weizenbaum-institut.de/index.php/wjds/article/view/3_3_1/113
[23] https://en.wikipedia.org/wiki/Ethics_of_artificial_intelligence
Slide 2: SNL AARP Alexa Silver edition YouTube Video
This parody of Alexa from May 13, 2017; first intro comments follow, you’ll be rolling on the floor laughing as you watch
It is “super-loud” for the hard of hearing, responds to “any name even remotely close to Alexa,” scans the room to help you find the phone in your hand, and boasts a special “Uh-huh” feature for listening to rambling stories.
It’s the “Echo Silver,” a partnership of Amazon and AARP, the “only smart speaker device designed specifically to be used by the greatest generation” — and it’s a pretty funny spoof. Here it is, as broadcast on Saturday Night Live May 13, 2017.
Or the direct YouTube link here:
Slide 3: Is there enough capital to deploy AI? Yes, but can it be powered?
Here is the link to the Goldman Sachs reference: https://www.goldmansachs.com/what-we-do/investment-banking/insights/articles/powering-the-ai-era?chl=ps&plt=go&cid=22699750214&agp=183774187120&kid=ai%20capital%20expenditure&mtype=p&gclid=Cj0KCQjw5onGBhDeARIsAFK6QJYrijF9TFvOTOKWVagNo1wSTja8aG9j21wOdWB4pN
Slide 4: Why is human knowledge accellerating?
Reference: https://bigthink.com/the-future/mass-flourishing-mark-mills/
Slide 5: Amount of Human knowledge is increasing at amazing speeds
Slide 6: How long till inventions are used by 25% of US population?
Slide 7: New inventions cause humans to lose necessary skills
Calculators? No Socrates lamented the written word as making it easy to show wisdom (without understanding).
Many lament the impact of technology feeling we use new ways as crutches and loose key knowledge and skills
Socrates felt the same about writing as compared to the discussions he favored in teaching others.
Slide 8: AI what is it?
What is AI?
Slide 9: AI 3 laws of robotics, slave or equal?
The play, written by Karel Čapek, describes the creation of enslaved synthetic humans, or robots—a term derived from robota, the Czech word for “forced labor.” Čapek’s robots, originally made to serve their human masters, gained consciousness and rebelled, soon killing all humans on Earth. In the play’s final scene, the robots reveal that they possess emotions just like we do, and the audience is left wondering whether they would also achieve the ability to reproduce—the only thing still separating robots from humans.
Nishimura would later explain in a newspaper article, he was troubled because the play convincingly portrayed “the emergence of a perverse world in which humans become subordinate to artificial humans.” Gakutensoku means “learning from the rules of nature.” Learn more at; https://www.nippon.com/en/views/b00906/
Isaac Asimov's laws were a direct response to the common trope of destructive robots in earlier science fiction. He developed them to provide a set of guiding principles that would keep robots under human control
Driven to avoid the "Frankenstein theme" of robots destroying their creators, Asimov created these laws to control artificial beings. The laws, designed to be the most fundamental programming for any robot
For relevance today, review https://sixdegreesofrobotics.substack.com/p/exploring-isaac-asimovs-three-laws
Classic ethical dilemma that illustrates the limitations of Asimov's laws is the trolley problem. First law of robotics conflicts with either choice.
Slide 10: AI history 2
1981: The Japanese government allocated $850 million (over $2 billion dollars in today’s money) to the Fifth Generation Computer project. Their aim was to create computers that could translate, converse in human language, and express reasoning on a human level.
Slide 11: AI History 3
Slide 12: AI History 4
Slide 13: Recent trends show weakening consumer economy but overall AI ramp still keeping US economy on track.
Nvidia's Q2 FY2026 revenue was $46.7 billion, with the Data Center segment accounting for the vast majority, around $41.1 billion, making it the primary driver of the company's growth and essentially representing its AI business. This figure represents a 56% year-over-year increase from the previous year's second quarter and shows continued demand for Nvidia's AI chips, despite the period being the slowest for overall growth in the last nine quarters of strong AI-driven expansion. [1, 2, 3]
Total Revenue (Q2 FY2026): $46.7 billion
Data Center Revenue: Approximately $41.1 billion
Gaming Revenue: Approximately $4 billion
Key Insights
AI is the Core: The Data Center segment, which includes AI chips, is the overwhelming driver of Nvidia's revenue. [1, 4]
Strong Year-Over-Year Growth: Revenue grew by 56% compared to the second quarter of the previous year. [2, 3]
Slowing Growth Rate: While still strong, the growth in the second quarter was the slowest it had been in the nine quarters since the current AI boom began, possibly due to factors like the slowdown in China and the impending launch of new Blackwell chips. [3, 5]
China Sales Absence: No shipments of Nvidia's high-end H20 chips for the Chinese market occurred during this quarter. [2, 3]
Future Outlook: The company expects to ship new Blackwell data center chips in the fourth quarter, which could further fuel future growth. [5]
AI responses may include mistakes.
[2] https://nvidianews.nvidia.com/news/nvidia-announces-financial-results-for-second-quarter-fiscal-2026
[3] https://www.cnbc.com/2025/08/27/nvidia-nvda-earnings-report-q2-2026.html
[4]
Not all images can be exported from Search.
Slide 14: The knowledge economy by valuations
Note ExxonMobil has the smallest valuation of this list of companies. This showes the impact of technology as the new wave of growth is knowledge driving rather than oil driven
Slide 15: AI applications, show me the money!
Slide 16: Arial Drones as wingman
https://dingeraviation.net/bf109/bf109tactics.html Luftwaffe fighter tactic
• Remember Messerschmitt Bf 109: formation of fighter tactics for fast metal monoplane aircraft that appeared at the end of the 1930s
•Aircraft fly far enough apart there is less chance of collision. The
•Leader looks forward while wingman's attention is to the rear, making sure an enemy aircraft does not sneak up from behind
•Best form of fighter attack is out of the sun, where an aircraft is very nearly impossible to see against the glare.
To guard against this the pair would position themselves so that each had a clear view of the sun, unobstructed by the other.
•Schwarm is two pairs acting together; aircraft flying wide apart (harder to see than tight grouping). Leaders looking forward, wingmen concentrating on rear
https://www.graphicnews.com/en/pages/41103/military-boeing-loyal-wingman-drone
March 5, 2021 - “Loyal Wingman”, a joint project between Boeing and Australia’s military, is a combat drone capable of flying alongside manned aircraft, offering fighter jet-like performance at a much lower cost.
The “Loyal Wingman” – the first military aircraft designed and manufactured in Australia in over 50 years – made its first flight on Feb 27.
The aircraft is 11.6m long, has a 2,000 nautical mile (3,700km) range, and a nose that can be outfitted with various payloads. The aircraft can also carry weapons and act as a shield to help protect more expensive manned fighter jets, according to its developers.
Boeing has decided to the Loyal Wingman as the basis for its $400 million U.S. Air Force Skyborg prototype, an executive at the plane maker said on Mar 2.
As part of the Skyborg programme, the U.S. Air Force hopes to build a family of low-cost, attritable drones that can be reused, but are cheap enough that losses in combat can be financially and operationally tolerated.
The project is meant to produce a family of uncrewed aerial systems that can move into contested spaces and conduct aerial missions that might be too dangerous for human pilots to perform.
Slide 17: Naval Drones as well
Slide 18: Self Driving Vehicles - Fully autonomous driving isn’t quite ready for prime time.
Liability issues and risks need to be addressed.
Developing news:
The Tesla class action lawsuit alleges that defendants throughout the Class Period made false and/or misleading statements and/or failed to disclose that: (i) Tesla overstated the effectiveness of its autonomous driving technology; (ii) there was thus a significant risk that Tesla’s autonomous driving vehicles, including the Robotaxi, would operate dangerously and/or in violation of traffic laws; (iii) the above increased the likelihood that Tesla would become subject to heightened regulatory scrutiny; and (iv) accordingly, Tesla’s business and/or financial prospects were overstated.
The Tesla class action lawsuit further alleges that on June 23, 2025, Bloomberg published an article entitled “Tesla Robotaxi Videos Show Speeding, Driving Into Wrong Lane,” which reported that “Tesla Inc.’s self-driving taxis appeared to violate traffic laws during the company’s first day offering paid rides, with one customer capturing footage of a left turn gone wrong and others traveling in cars that exceeded posted speed limits.” That same day, the complaint also alleges, in an article entitled “Tesla Robotaxi Incidents Draw Scrutiny From US Safety Agency,” Bloomberg reported that the U.S. National Highway Traffic Safety Administration (“NHTSA”) had contacted Tesla regarding the foregoing incidents, noting that the NHTSA “is aware of the incidents that were captured in videos posted on social media and is gathering additional information from the company.” Further, the Bloomberg article quoted a statement released by the agency that “‘[f]ollowing an assessment of those reports and other relevant information, NHTSA will take any necessary actions to protect road safety,’” according to the Tesla class action lawsuit.
One of my favorite editors and longtime friend is Junko Yoshido. Her current endeavor highlight many of the key concerns about the state of driverless vehicles.
Another driverless vehicle expert is Missy Cummings, professor at George Mason University. https://www.linkedin.com/in/missy-cummings-7681588b/
Dr. Cummings is considered an international expert and leader in the field of autonomous systems by such agencies as the US Department of Defense, the World Economic Forum, New America Foundation, BBC, the American Bar Association, Wired, the AP Press, NPR, the Economist, Scientific American, Time, NOVA, PBS, NBC, PBS, Fox News, and many others. She has written several opinion editorials for agencies like CNN and Wired and has made several radio, podcast and TV outreach appearances. She recently served as the Senior Safety Advisor to the National Highway Traffic Safety Administration.Dr. Cummings is considered an international expert and leader in the field of autonomous systems by such agencies as the US Department of Defense, the World Economic Forum, New America Foundation, BBC, the American Bar Association, Wired, the AP Press, NPR, the Economist, Scientific American, Time, NOVA, PBS, NBC, PBS, Fox News, and many others. She has written several opinion editorials for agencies like CNN and Wired and has made several radio, podcast and TV outreach appearances. She recently served as the Senior Safety Advisor to the National Highway Traffic Safety Administration.
https://www.engineergirl.org/2923/Missy-Cummings
Slide 19: Robots ready for manufacturing, entry costs of $25,000 and up encourage deployment and workflow optimization for 24x7 operations.
Video links
Boston Scientific robots dancing to Do You Love me.
Tesla Optimus Robot dancing balley and rap to show dexterity and balance: https://www.facebook.com/reel/1463281661498111
•Digit from Agility robotics demonstrates the value of human physical apprearance in ease of application of robotics to material handling tasks optimized for humans initially.
•Automated AI robotic Warehouse where 46 robots and a handful of humans manage online fulfillment.
Slide 20: Robotics can change medical applications and improve care at lower cost with better end results
Robotic Surgery clips
How robots learn to do stitches and autocorrect for problems or interference.
Surgical outlook is better for patient. They has smaller incisions, less bleeding and go home sooner.
Slide 21: Lots of AI choices, play with a few, find those you like the most.
Free is good place to start, research how your data is protected - - no worse than Google search, but not the place for confidential info (passwords, key trade secrets, customer lists, etc).
Most of these AI tools have a “free” use category. Do a little research asking Google Search of even these AI’s to search and provide info on security, data privacey and confidentiality of your inputs and your work product before doinging significantl development work and/or sharing data you don’t want others to see.
My usage has mostly been with copilot, perplexity and gemini. Every so ofter you’e used your free queries for the day and can’t do another search till the next day. If you’re on a deadline, then copy the work you’ve done so far and use another AI engine to continue by submitting the file of work today for input and also the next question you have and want it to work on.
https://www.anthropic.com/claude
https://www.llama.com — opensource can download tools to us (more for DIY rather than access AI online)
https://cohere.com/command appears to be aimed at enterprise users
https://stability.ai Stability AI models are free for non-commercial use and for commercial use if your annual revenue is below $1 million.
https://www.midjourney.com/home All midjourney plans are paid subscriptions; start at $10/mo.
https://www.deepl.com/en/translator
https://character.ai just looked and this is interesting!
https://app.runwayml.com Upgrade to a paid plan to use Text to Video with integrated audio and dialogue
Slide 22 AI tools are useful. Practice to learn how to get what you need efficiently.
•Organize family schedule and tasks
–https://ollie.ai/2024/07/22/how-family-ai-is-becoming-moms-secret-weapon-against-mental-load/
Slide 23 AI doing up to 50% of the work at Salesforce
Tax policies that allowed expensing of technical development efforts combined with low cost of money, where rates, equal or less to inflation, encouraged companies to hire excessively as money/people were virtually free even with modest results.
As the cost of money increases to 8% or so and tax advantages weaken, now folks questioning “what did you do for me today” and determining how to sort the best from subpar or even average employees for cost reasons.
FOMO (fear of missing out) is driving businesses to highlight how much they use AI (whether critical to success or just flailing and trying).
As such cutting bloated staffing is highlighted as redeploying into AI so shareholders, executive and bosses in general think their company is doing well in AI vs others. We shall see as the results come in who the winners really are.
Salesforce CEO Marc Benioff revealed the layoffs during an interview published Friday Aug 29, 2025 on The Logan Bartlett Show podcast.
“I’ve reduced it from 9,000 heads to about 5,000, because I need less heads,” Benioff said while discussing the impact of AI on Salesforce operations.
AI doing up to 50% of the work at Salesforce
“Because of the benefits and efficiencies of Agentforce, we’ve seen the number of support cases we handle decline and we no longer need to actively backfill support engineer roles,” Salesforce said in a statement Tuesday to NBC Bay Area.
Went on to note, that when distributing leads to human salesforce, sales people choose the best leads to work. Benioff noted 50% of leads aren’t called. With AI agents, all leads are contacted. And like most quick data choices, there will still be some good quality leads in those initially discounted.
Sales cold calls, “smile and dial,” typically are a 1-3% hit rate. A sales person needs to make 100 contacts in an hour or so, then spend the rest of their day working those good leads.
Slide 24: What happens to the workforce? Become a knowledge worker for success!
Manufacturing jobs are not coming back to US as a major portion of the economy. Insufficient engineering talent is available to staff the requirements of setting up production level manufacturing. Direct labor STEM skills are also weak, challenging one to find adequate numbers of capable workers with technical skills.
Trades jobs have similar problems in attracting workers with sufficient STEM and technical skills.
Service industry under assault for entry jobs
Source: https://www.buzzmaven.com/old-engineer-hammer-2/
Slide 25: two knowledge workers use of AI
https:goupslope.com website created with AI tools as implementer and person as “architect” content expert and know what interactions looks like.
https://www.laticrete.com LATICRETE is a family-owned, customer-focused, worldwide building material manufacturer and supplier: 100 Countries, 3 Generations, 1 Mission We aim to be the leading construction brand trusted globally for high-performance tile and stone installation systems and building finishing solutions.
Slide 26: Call to Action
Now put your newly used knowledge to work and engage with AI!
Closing Comments
Thank you for following my; feel free to share this content with others. The better we understand AI, the better we can take advantage of the tremendous capabilities to help us bring order as we seek new knowledge in a rapidly changing world.
Live long and prosper!