Category: Artificial Intelligence

On this final day of the 12 Days of AI, participants are asked to reflect a bit on the activities from the past eleven inquiry prompts. I’ve been an outsider of sorts in this project, since I never joined the Microsoft Teams set up for discussion and I imagine — based on the prompts — that the people who are participating are university people in Europe, not elementary teachers in the US, like me.

I tagged along because I am curious. Though wary of the power and potential of the some of the Generative AI tools out there, I admit that, as an educator and writer and artists, I wonder about how we might find ways to harness these tools for creativity. So I played along, and explored some platforms I already knew and some I didn’t know, and tried to be thoughtful about ethics, data and learning as I went along.

I enjoyed the Mapify platform for making AI Mind Maps. I returned to the Holly AI site to explore sound, which was a strange experience and I am not sure the practicality of it. I used a few video AI generating sites, and while I could see all of the flaws, I could also recognize the potential and the potential harm as these sites get better and smoother, and the AI deep fakes and fake people get more realistic. I am curious about Miro, a site I didn’t tap because it seemed like it was available already to the university folks but it seems to more multi-media Generative AI than most. I wondered how the university people were using it, to be honest.

The final activity is to create an artifact of the 12 days of learning, maybe by moving across platforms.

I decided to ask Mapify to create a map of learning for someone new to Generative AI, who wants to learn slowly, and end with a final project. It did. I like how it ended in a creative art project.

I then took that concept into both the Holly audio transformer and an AI Voice site, added some text to be spoken, and then wove those audio files together. For the image, I took the Mind Map and used it as a visual reference in Adobe Firefly, and asked it to create an image of someone new to AI, in exploration.

The result is odd, and the seamlessness of crossing AI platforms is not there yet.

My next steps? Go back into the accounts I used for 12 Days, and delete the ones I know I will never use again.

Peace (splattering paint),
Kevin

La Commedia umana – The Human Comedy flickr photo by JuanGalvez68 shared under a Creative Commons (BY-SA) license

On this eleventh day of the 12 Days of AI, the theme was to look at asynchronous collaboration with Mira, an AI platform that I don’t have access to (but the university hosting the 12 Days, does — I am taking part of in the 12 Days as an outsider) so I went into Claude, instead, to explore the concept through conversation with Claude, asking questions about the pros and cons of live interactions with Generative AI. (Mira can produce sticky notes, artwork, graphs, etc. so my explorations here were mostly limited to text responses).

Pros Of Collab (via Claude)

• Enhanced Productivity and Efficiency Asynchronous AI collaboration allows humans to work on complex tasks without being constrained by real-time interactions. Users can initiate a task, receive detailed outputs, and then refine or build upon the AI’s work at their own pace. This approach enables more thoughtful, iterative problem-solving without the pressure of immediate responses.
• Flexible Knowledge Expansion Unlike synchronous interactions, asynchronous collaboration gives humans time to deeply reflect on AI-generated content. They can carefully review suggestions, research additional context, and gradually develop more nuanced understanding. This method supports deeper learning and more comprehensive exploration of complex topics.
• Reduced Cognitive Load By breaking down complex tasks into manageable chunks, asynchronous AI collaboration helps reduce mental fatigue. Humans can tackle different aspects of a project incrementally, leveraging AI to handle initial drafting, research, or analysis, and then focus their energy on high-value creative and strategic refinements.
• Comprehensive Document and Project Development Asynchronous modes are particularly powerful for long-form content creation. Whether it’s writing reports, developing software, creating educational materials, or designing complex documents, humans can gradually shape and refine AI-generated content, adding personal insights, context, and nuance over time.
• Persistent Context and Iterative Improvement Modern AI collaboration tools maintain context across interactions, allowing for continuous refinement of work. Users can return to previous drafts, request specific modifications, and build upon earlier AI-generated content without losing the thread of their original intent.
• Enhanced Creativity and Ideation Generative AI can serve as a powerful brainstorming partner in an asynchronous mode. It can quickly generate multiple perspectives, draft initial concepts, or provide creative prompts that humans can then critically evaluate, modify, and expand upon at their own pace.
• Skill Development and Learning Asynchronous AI collaboration provides a unique opportunity for skill development. Users can observe how AI approaches problems, learn from its initial drafts, and gradually develop more sophisticated prompting and collaboration techniques.
• Accessibility and Time Zone Independence Unlike synchronous interactions, asynchronous AI collaboration transcends time zone limitations. Users can work on projects whenever convenient, receiving sophisticated AI assistance that doesn’t require simultaneous availability.
• Risk Mitigation and Verification The slower pace of asynchronous collaboration allows for more thorough fact-checking, verification, and refinement. Humans have time to cross-reference AI-generated content, ensuring accuracy and reducing the risks of misinformation or hasty decision-making.
• Personalized and Contextual Outputs Through iterative interactions, users can progressively refine AI outputs to better match their specific context, writing style, or project requirements. This collaborative approach produces more tailored and precise results compared to one-off interactions.

My Observation: The “creativity and ideation” point is one that intrigues me, I think, and wonders how artists and creative people can best harness this technology to push art in new directions. Not by just asking AI to do something, to make something, but to move to the edges of the possible, and then help the artist go a few steps further.

Cons of Collab (via Claude)

• Potential for Misinterpretation Without real-time dialogue, there’s a higher risk of misunderstanding the original intent. Nuanced requirements or complex context can be lost or misinterpreted between interaction cycles, leading to progressively less relevant AI outputs.
• Increased Cognitive Overhead Paradoxically, while aiming to reduce cognitive load, asynchronous collaboration can actually increase mental effort. Users must carefully craft initial prompts, review outputs critically, and manage multiple iterations of refinement, which can be more mentally taxing than direct, synchronous communication.
• Context Drift Over multiple interactions, the original project goal can subtly shift or become diluted. Without immediate clarification, AI might progressively move further from the initial objective, creating outputs that become increasingly disconnected from the user’s original vision.
• Reduced Spontaneity and Dynamic Interaction Real-time dialogue allows for immediate clarification, spontaneous idea generation, and dynamic problem-solving. Asynchronous modes lose this conversational fluidity, making the collaboration feel more mechanical and less organic.
• Delayed Problem Resolution Complex problems that would benefit from immediate back-and-forth dialogue get stretched out over longer periods. What might take 15 minutes in a synchronous conversation could extend to hours or days in an asynchronous mode.
• Increased Risk of AI Hallucinations Without immediate human intervention, AI might generate more elaborate or convincing but factually incorrect information. The lack of real-time fact-checking increases the potential for sophisticated but inaccurate content.
• Emotional and Creative Disconnection Asynchronous collaboration can feel impersonal. The iterative, delayed nature removes the energetic exchange of ideas, potentially reducing the emotional engagement and creative spark that comes from real-time brainstorming.
• Dependency and Skill Atrophy Overreliance on asynchronous AI collaboration might lead to diminished personal problem-solving skills. Users might become too dependent on AI-generated content, potentially reducing their own critical thinking and creative capabilities.
• Complex Workflow Management Tracking changes, managing multiple versions, and maintaining a coherent development process becomes more challenging. Users must carefully manage versions, remember context across interactions, and maintain a clear mental model of the project’s evolution.
• Technological and Interface Limitations Not all AI collaboration tools are equally equipped for asynchronous work. Some platforms have limited version tracking, poor context retention, or clunky interfaces that make iterative collaboration frustrating.
• Ethical and Accountability Challenges In asynchronous modes, the lines of intellectual ownership and contribution become blurrier. It’s less clear who is truly responsible for the final output – the human, the AI, or some hybrid collaboration.
• Performance and Cost Considerations Multiple interactions and iterations can potentially increase computational costs and response times, especially if working with more advanced AI models that require significant processing power.

My Observation: So many of these are on my own radar, but the point about “dependency and skill atrophy” and the over-reliance on AI was interesting, as you can see a version of that happening with the advent of smart phones and GPS mapping, and how our reliance on our technology has changed the way we learn, use and retain new information in our heads. If everything is infused with Generative AI, will we still find ways to think through problems on our own and find original, creative solutions?

I also wondered about real-world examples of positive collaborations. Claude abides.

1. Medical Diagnostics and Research

• Radiologists use AI algorithms to analyze medical imaging, detecting potential tumors or abnormalities with greater accuracy and speed.
• Researchers at Stanford University developed an AI system that can detect skin cancer with accuracy comparable to dermatologists, serving as a screening tool that assists human experts rather than replacing them.
• Drug discovery teams use AI to predict molecular interactions and potential drug candidates, dramatically reducing the time and cost of developing new medications.

2. Scientific Research

• NASA uses AI to analyze vast amounts of astronomical data from telescopes, helping scientists identify potentially habitable exoplanets and unusual cosmic phenomena.
• Climate scientists leverage AI to process complex climate models, identifying patterns and making more accurate predictions about climate change impacts.
• Protein folding research saw a breakthrough with DeepMind’s AlphaFold, which helps researchers understand protein structures, accelerating biological and medical research.

3. Creative Industries

• Film and animation studios use AI for pre-visualization, helping artists quickly prototype complex scenes and visual effects.
• Musicians collaborate with AI tools to generate unique musical compositions, explore new melodic patterns, and overcome creative blocks.
• Graphic designers use AI-powered tools to generate initial design concepts, which they then refine and customize.

4. Education

• Personalized learning platforms use AI to adapt curriculum to individual student needs, providing tailored learning experiences.
• Language learning apps like Duolingo use AI to create adaptive learning paths, helping humans learn languages more efficiently.
• Teachers use AI tools to assist with grading, identify students who might need additional support, and develop more personalized teaching strategies.

5. Environmental Conservation

• Conservationists use AI-powered drone and satellite imagery to track wildlife populations, monitor deforestation, and detect illegal poaching activities.
• Ocean researchers employ AI to analyze marine ecosystem data, tracking marine life migrations and understanding climate change impacts.

6. Agricultural Innovation

• Farmers use AI-powered systems to optimize crop yields, predict pest infestations, and make more precise irrigation decisions.
• Precision agriculture technologies use AI and satellite imagery to provide detailed insights about crop health, soil conditions, and potential improvements.

7. Accessibility Technologies

• AI-powered translation tools help break down language barriers, assisting people in real-time communication across different languages.
• Assistive technologies use AI to help people with disabilities, such as text-to-speech systems for visually impaired individuals or communication aids for those with speech impairments.

8. Cybersecurity

• Security teams use AI to detect potential cyber threats, analyze network traffic, and identify unusual patterns that might indicate a security breach.
• AI helps human cybersecurity experts by providing real-time threat detection and predictive analysis of potential vulnerabilities.

9. Emergency Response and Disaster Management

• AI helps predict natural disasters and optimize emergency response strategies.
• During the COVID-19 pandemic, AI was used to track virus spread, assist in vaccine development, and help health authorities make more informed decisions.

10. Business and Customer Service

• Customer service teams use AI chatbots for initial screening and handling routine inquiries, allowing human agents to focus on more complex customer needs.
• Marketing teams leverage AI for data analysis, helping to develop more targeted and effective marketing strategies.

It was this last section, on real-world applications, that makes me hopeful that the advancements in AI (did you read that Google says it has begun Quantum Computing? That’s huge!) could have a positive impact on our world, through fields of health and science, in particular. But of course, there will need to be more guardrails.

Peace (pondering),
Kevin

On this tenth day of the 12 Days of AI, we are exploring the bias that comes within AI systems, either intentionally or unintentionally. All systems have inherent bias because of the data sets that work beneath the interface. Some AI sites do a better job at countering bias than others.

The post reminds us:

“AI generative output is not creativity, but a statistical variation without intent and meaning. AI is a statistical artist. It introduces variations into its work, but its “choices” are driven on patterns and probabilities rather than deep understanding of lived experiences and emotions. AI generative outputs, while devoid of the intent and meaning characteristic of human creativity, possess significant potential to speed, support, and extend the creative process.”

The prompt was to go into an AI image generator, and use a prompt of a person in a setting, doing some action (they frame it as a scene from a movie) and look to see if there were any apparent bias in the results. I used Adobe Firefly and asked it to generate a “stern-looking school principal talking to an elementary teacher about student work.”

Is there bias in the two collections of results I received?

Out of the eight images, it seems as if five (and maybe six? Depends how you look at who is who) show a white male in the role of principal, and of the eight, only two images show non-white adults being generated. Two women seem to be in the administrative role. I didn’t have enough credits in Firefly to keep running the experiment a few more times, which would have been ideal. Would I have kept finding similar genders and races of the administrators?

What about if I refined my prompt, adding loaded terms like “urban school” or “struggling teacher” or “disruptive student”? How would other platforms perform with the same prompt? Would there be some gender/racial stereotypes into who is an administrator and who is an educator?

Peace (in explorations),
Kevin

How Does Digital Technology Affect You? flickr photo by schopie1 shared under a Creative Commons (BY-SA) license

On this ninth day of the 12 Days of AI, we are looking at Deep Fake technology, examining both the possibilities and the concerns of how advancements in video production powered by AI makes for possibilities. We are asked to create our own Deep Fake video through a site called HeyGen, with a script created by Generative AI. I used Claude, and asked it to generate a script about jazz pianist Thelonius Monk.

I wasn’t necessarily all that impressed with the results, to be honest (but she seems pretty happy in her delivery). Still, you can see the potential, and maybe a paid account gives more flexibility and tools for voice, etc.

I followed the steps, but choices are pretty limited with a free account. I chose a host, and a voice, added the script, and al little text and background image. I wouldn’t quite put what I did under the Deep Fake category — given all the synthetic media I used. I think of Deep Fakes more connected to real people’s image and voice and video.

As it turns out, I have plans to chat about Deep Fakes with my students today, as part of a larger unit about technology in their lives. So I appreciated this chart, shared off the main post today:

Since the 12 Days is run by university folks, one of the ethical questions here is whether these AI videos could play a role in the delivery of education at an institution. Lord, I hope not. I can’t imagine sitting through a course with an AI Deep Fake Teacher, but I bet somewhere, some administrators are definitely imagining it (and thinking of the cost savings).

Peace (real, if we imagine it),
Kevin

cybermacs_huge_title_text_VOICE_four_cute_and_retro_robots_are__0788eb40-f0de-40b9-9310-af0b084dca2c flickr photo by CyberMacs shared under a Creative Commons (BY-NC-SA) license

It’s the eighth day of the 12 Days of AI project, and today, the prompt has us looking at the affordances, limitations and possibilities of AI-generated video content. The site we are exploring is Synthesia, which is apparently geared towards the business world, and not education, and as I explored the site with a prompt to create a learning video, I had to navigate elements of the business world lexicon. I also realized, mid-way through, that the site works best on Chrome or Edge, and not Firefox, but I was already knee-deep in it, so I kept going.

I decided to ask Synthesia to work on a video for me about a topic I am teaching today, in fact, as part of our vocabulary lesson. The non-fiction text for today’s lesson is all about silk works and the making of silk. So I uploaded the student text as PDF and asked it create a video about silk worms for elementary students.

Here is the AI video it created:

 

I recently tried Google’s AI Video platform (now embedded inside my Google Drive) and it was terrible on this very same topic. The information was not quite correct, and the flow of the interface and wording was so disjointed I abandoned my efforts to make it work. Synthesia, on the other hand, is smooth and seemed to do a pretty decent job of integrating the information I uploaded (about silk worms) into the video presentation.

I only used the free version, so I could not do much tinkering with the template (can you tell this is a site aimed at businesses?) and I was put off a bit at the end, as the AI woman presenter wraps things up with a sultry voice out of place with the rest of the presentation. I cringed, a bit. At times throughout the video, you can hear the voice being stitched together and some of her expressions seem odd. But still, this is a big improvement from where things were just a year ago in AI video production, I would say.

I also realize that I have been seeing more of these same kinds of AI talking head videos in the online world, and I imagine some businesses are going be all-in on this, creating AI training videos and AI marketing videos and more, without need for human editors or media employees, and how sad is that? (I have to sons who are in the film-making field and another in public relations and I worry mightily for their futures).

So, will I share this with my sixth graders? Maybe. We’ve been doing work around digital technology in their lives and we just did some work the other day exploring the role of Generative AI in the world. They might be intrigued by this, and perhaps bring a critical lens to the production value here.

Peace (in the vid),
Kevin

2022-02-11_21-13-24_ILCE-7C_DSCAG3501_DxO flickr photo by Miguel Discart & Kiri Karma shared under a Creative Commons (BY-SA) license

Oh, the irony of ironies – I used Google’s NotebookLM to help me with this task of looking at the environmental costs of AI through the lens of three different articles. What a slacker I am! And in looking at the environmental impact, I added to it on this seventh day of the 12 Days of AI.

BUT the NotebookLM AI was able to summarize and collate and thematically string together facts and insights from across the three articles in a very efficient way. The articles were:

• Via Kleinman Center for Energy Policy: https://kleinmanenergy.upenn.edu/commentary/podcast/why-ai-consumes-so-much-energy-and-what-might-be-done-about-it/
• Via Tech Target: https://www.techtarget.com/searchenterpriseai/feature/Energy-consumption-of-AI-poses-environmental-problems
• via Tom’s Guide: https://www.tomsguide.com/ai/chatgpt-energy-emergency-heres-how-much-electricity-openai-and-others-are-sucking-up-per-week

Here is the summary by AI:

Two reports reveal the substantial environmental impact of AI, particularly large language models like ChatGPT. The energy consumption for a single ChatGPT query is significantly higher than a standard Google search, leading to increased demand for data centers powered by energy-intensive sources like natural gas. This surge in demand also results in higher water usage for cooling servers, with varying consumption levels depending on geographic location. Experts predict that the problem could worsen tenfold by 2030 unless efficiency improvements are made.

And I found this factoid alarming and fascinating:

A single ChatGPT-4 Query can use the equivalent electricity of powering a home for 20 days and consume up to 3 bottles of water for cooling.

I remember looking into this topic for ETMOOC2 back in April, when it was clear that the advances in AI was leading to a surge in electricity and cooling. This diagram from Environmental Science & Technology is still worth examining.

And of course, I had to use the AI podcast feature in NotebookLM (which I notice has some recent updates — an option to give the AI audio tool guiding points to talk about). I asked it to examine how much of an impact on the environment went in to making ITS own podcast. At the 7:27 mark, it does examine its own footprint, if only briefly.

Here is the 15 minute podcast.

 

I’ll end by sharing what my AI inquiry says can be done to mitigate the impact on the planet.

Several strategies can help mitigate AI’s energy footprint:

• Improving model efficiency: Developing AI models that require less data and computational power for training and inference is crucial. This can involve techniques like model compression and quantization to reduce energy consumption during inference.
• Optimizing data center infrastructure: Investing in energy-efficient data center infrastructure, including advanced cooling systems and intelligent power distribution, can significantly reduce energy waste. Additionally, strategically locating data centers near renewable energy sources or in areas with lower environmental impact can help minimize reliance on fossil fuels.
• Responsible data management: Focusing on generating high-quality data instead of accumulating massive amounts of low-quality data can reduce storage needs and associated energy consumption. Regularly deleting unused data and promoting data reuse and recycling can further contribute to energy savings.

So I guess I became part of the problem by looking into the problem with AI. Sigh.

Peace (and pondering the planet),
Kevin

PS — and then later in the morning, I saw a post shared on BlueSky by Jennifer Elemen with this presentation link that connects nicely to this inquiry: https://docs.google.com/presentation/d/1UEpB2Hp-Ds_vZQHmwZPM63qN8cuhyZURBnp82ahF3yA/edit#slide=id.g303c8e51bcb_0_0

Question flickr photo by ryanvanetten shared under a Creative Commons (BY-SA) license

Here is the fourth day in the 12 Days of AI that I am engaged in, exploring a variety of approaches and tools, with reflection. The prompt today was to create a multiple choice set of questions, with thoughts on being specific and then comparing across platforms. I chose the topic of bird songs and music composition, as I recently did a whole inquiry project on that topic and AI.

Here is my prompt for Generative AI: Create three multiple choice questions on the topic of how music composers use bird song as inspiration for creating music

---

This is what Claude gave me:

1. Which famous classical composer was known for precisely transcribing bird songs and incorporating them almost exactly into his compositions?

A) Wolfgang Amadeus Mozart
B) Olivier Messiaen
C) Ludwig van Beethoven
D) Johann Sebastian Bach

2. In Olivier Messiaen’s landmark work “Quartet for the End of Time,” how did he specifically utilize bird songs?

A) He recorded actual bird sounds and played them during the performance
B) He meticulously transcribed the precise musical pitches and rhythms of bird calls into musical notation
C) He used bird-themed titles but did not directly incorporate bird songs
D) He used electronic synthesizers to simulate bird sounds

3. What musical technique do composers often use when drawing inspiration from bird songs?

A) Completely avoiding any rhythmic patterns
B) Mimicking bird calls through instrumental techniques like trilling, warbling, and unusual pitch variations
C) Only using bird songs in purely electronic music
D) Restricting bird song inspiration to percussion instruments

Answers:

1. B) Olivier Messiaen
2. B) He meticulously transcribed the precise musical pitches and rhythms of bird calls into musical notation
3. B) Mimicking bird calls through instrumental techniques like trilling, warbling, and unusual pitch variations

The questions highlight the fascinating intersection of ornithology and musical composition, particularly focusing on how composers like Olivier Messiaen have been deeply inspired by the complex musical qualities of bird songs.

---

Here is what Google’s Gemini gave me:

Here are three multiple-choice questions about how music composers use bird song as inspiration:

Question 1: What is a common technique used by composers to incorporate bird song into their music?

• A) Directly transcribing the exact pitch and rhythm of the bird song.
• B) Using the bird song as a starting point for melodic and harmonic ideas.
• C) Sampling the bird song and incorporating it into electronic music.
• D) All of the above.

Question 2: How does bird song influence the emotional quality of a musical composition?

• A) Bird song can evoke feelings of peace and tranquility.
• B) Bird song can convey a sense of joy and exuberance.
• C) Bird song can create a sense of mystery and intrigue.
• D) All of the above.

Question 3: Which famous composer is known for their extensive use of bird song in their music?

• A) Wolfgang Amadeus Mozart
• B) Johann Sebastian Bach
• C) Olivier Messiaen
• D) Ludwig van Beethoven

---

Here is what ChatGPT gave me:

1. How do composers typically incorporate bird songs into their music?

A) By mimicking the exact pitch and rhythm of the bird songs
B) By using bird calls as a rhythmic foundation or melody in their compositions
C) By turning bird songs into operatic vocal performances
D) By using bird song recordings as background noise without altering them

Answer: B) By using bird calls as a rhythmic foundation or melody in their compositions

---

2. Which famous composer is known for integrating bird songs into his works, particularly in pieces like “The Carnival of the Animals”?

A) Claude Debussy
B) Ludwig van Beethoven
C) Pierre Boulez
D) Camille Saint-Saëns

Answer: D) Camille Saint-Saëns

---

3. What is one reason why composers are inspired by bird songs in their music?

A) Bird songs are always structured in classical time signatures
B) The unpredictable and varied nature of bird calls can inspire unique musical phrasing
C) Bird songs are easy to replicate on traditional instruments
D) Bird songs are often associated with harmony and major scales

Answer: B) The unpredictable and varied nature of bird calls can inspire unique musical phrasing

---

Reflection: Each of the three Generative AI sites (Claude, Gemini, ChatGPT) presented me with something slightly different but with some common ideas, and I could see choosing questions from each to create an interesting quick quiz on this topic. I happen to like the generated questions from each site that directly connects the making of music to the listening to bird song.

Peace (singing it),
Kevin

PS — I went a step further and asked Google Gemini INSIDE of Google Forms to create a quiz, based on the same question. Here it is, if you want to see how well you will do: https://forms.gle/WucmRiL8Pm3oXAcQ8