Pedro Arrupe, S.J. Center for Business Ethics’ Post

LLMs are super useful for increased productivity subject to creativity 😎 Thought of Evaluating My draft PhD Proposal using various evaluation matrices and perspectives. As a conclusion, thought of using the below Prompt in (1) ChatGPT, (2) Gemini and (3) Claude respectively --- Prompt Used --- Use Below Response Template Acceptance Status: [Accepted / Conditionally Accepted / Rejected] Score: [Points out of 100] Rational: [A single Sentence] --- End of Prompt --- Check out the responses of each LLM in the attached image 😎 !!!

ChatGPT linked to declining academic performance and memory loss in new study https://bit.ly/3IPQPhu

Interesting read! What are everyone's thoughts on this resource as it continues to expand?

"Alice has N brothers and she also has M sisters. How many sisters does Alice’s brother have?" Most state-of-the-art LLMs struggled with this basic question, except ChatGPT-4o, which answered correctly. New research highlights that LLMs often fail simple common-sense tasks solvable by humans, giving incorrect answers and nonsensical justifications. Standard interventions like enhanced prompting or multi-step re-evaluation don't fix these errors. The study urges re-assessing LLM capabilities and developing better benchmarks to detect reasoning flaws. Explore the findings and access the experiment code at https://lnkd.in/dMP9-WZf.

I see now why we need to replace the useful idiots in the Ivy league universities with AI

ChatGPT as a Patient Resource? In this study ChatGPT3.5 was able to answer 6/10 FAQ about UCL reconstruction with excellent or satisfactory performance… Open access ASMAR article: https://ow.ly/6BxG50SnMhA

Hi ed researchers! Just sharing some of the work colleagues from Center for Advanced Study of Teaching and Learning have recently made available via preprint (Jamie DeCoster et al.). Check these out: A paper on using ChatGPT for ed research (https://lnkd.in/gdC_UXsp) and another on data validation to support the trustworthiness of our work (https://lnkd.in/gTNdM2RE).

Using #AI to assess is a slippery slope. Don’t remove the human from the process. For me, using AI for assessment help Isn’t about decreasing workload. I still review the assessment. It’s about speed and development. With the help of of AI Students get (almost) instant feedback. Students can iterate assignments. I can grade the iterations. And they learn by reflecting and refining. Instead of the high-stakes of getting one shot They can improve as many times as they want. Isnt the goal to be developmental? That’s what AI helps us be as teachers. Don’t let it replace you. But don’t ignore it either. Your students will be the ones who suffer.

Reflecting on history and the intersection of technology and education, this image of protesting math teachers in 1988 advocating to ban calculators strikes a chord. Fast forward to 2023, and the sentiment hasn't strayed far. "We should ban ChatGPT," some say. Yet, amidst these calls for restriction, I ponder: "If you can't beat them, join them!" There's wisdom in embracing technological advancements. However, I resonate with the teachers' concerns. Every tool, including technology, has its pros and cons. Regulating technology usage in education could foster critical thinking and preserve the essence of individual thought. Technology should augment and simplify the human thought process, not overshadow it with a monitoring system that stifles originality. Let's strive for a balance where technology adds value while honoring the richness of human intellect. What are your thoughts? #Education #Technology #CriticalThinking

👋 Juliano Mologni - I had some deep shower thoughts today and came up with something I’m going to do. I think you’ll be interested based on how passionate you are with teaching others 🤓 I’m creating a WesWorld Tinkering Kit full of PCB madness—basically showing the “scary side” of EE concepts. 🎶 - Some original ideas by Wes + a little AI help, because why not? So far: 📞 Cross Talk – 2 signals, one heavily affected by cross talk, and the other clean. Maybe a speaker to audibly demonstrate cross talk. 💥 I Can’t Resist – Resistors for a satisfying boom with safety. 💥 No Cap – Capacitors overcharged and safely blown up. ⚡ Inductive Kickback – Relay or motor demo showing high-voltage spikes and protection with diodes. 🔥 Thermal Runaway – Transistor demo on the importance of heat dissipation. 💡 Voltage Divider Magic – Show voltage splitting with an adjustable LED brightness. 🚨 Reverse Polarity Protection – Diodes protecting against wrong power connections. 🌡 Overcurrent Protection – Fuses preventing circuit damage during current spikes. 🎛 RC Time Constant Fun – Explore time constants with adjustable resistor/capacitor circuits. 🌀 Magnetic Field Madness – Demonstrate electromagnetism with inductors and coils. ⚡ Grounding and Shielding Demo – EMI noise reduction with proper grounding. 📡 Impedance Mismatch Madness – Show signal reflection with impedance mismatches. 🛠️ Signal Integrity Tools – Visualize signal degradation using analyzers or scopes. 📏 Trace Width Experimentation – Adjust PCB trace widths to explore resistance and heat dissipation. 🛡️ Dielectric Effects – Demonstrate how dielectric materials affect signal speed and impedance. 🔄 Return Path Confusion – Show interrupted return paths and why they’re crucial in multi-layer PCBs. 🤖🤖🤖🤖🤖🤖🤖🤖🤖🤖🤖🤖🤖 ChatGPT 🤖 - This list will make the WesWorld Tinkering Kit both fun and educational, bringing practical EE concepts to life! Wes Lorenzini - Thanks friend ☺️ 🤖🤖🤖🤖🤖🤖🤖🤖🤖🤖🤖🤖🤖 PS: I play with Snap Circuits but feel there’s room for more love.