Artificial Intelligence/Agents/LLMs/and Related

??Well this is something else.

GPT-4 passes basically every exam. And doesn’t just pass…
The Bar Exam: 90%
LSAT: 88%
GRE Quantitative: 80%, Verbal: 99%
Every AP, the SAT… pic.twitter.com/zQW3k6uM6Z

— Ethan Mollick (@emollick) March 14, 2023

Description of video:

Sal Khan walks through Khan Academy’s GPT-4 integration (not generally available yet). Folks can join the waitlist at Khanacademy.org. To learn more about Khanmigo, visit: khanacademy.org/khan-labs

We believe that AI has the potential to transform learning in a positive way, but we are also keenly aware of the risks. To test the possibilities, we’re inviting our district partners to opt in to Khan Labs, a new space for testing learning technology. We want to ensure that our work always puts the needs of students and teachers first, and we are focused on ensuring that the benefits of AI are shared equally across society. In addition to teachers and students, we’re inviting the general public to join a waitlist to test Khanmigo. Teachers, students and donors will be our partners on this learning journey, helping us test AI to see if we can harness it as a learning tool for all.

GPT-4 has arrived. It will blow ChatGPT out of the water. — from washingtonpost.com by Drew Harwell and Nitasha Tiku
The long-awaited tool, which can describe images in words, marks a huge leap forward for AI power — and another major shift for ethical norms

Introducing Our Virtual Volunteer Tool for People who are Blind or Have Low Vision, Powered by OpenAI’s GPT-4 — from bemyeyes.com
We are thrilled to announce Be My Eyes Virtual Volunteer™, the first-ever digital visual assistant powered by OpenAI’s new GPT-4 language model.

For example, [GPT-4] passes a simulated bar exam with a score around the top 10% of test takers; in contrast, GPT-3.5’s score was around the bottom 10%.

Source

Here are some incredible things people are already doing with GPT-4

It’s been less than 3.5 hours

? A thread

— Linus (???) (@LinusEkenstam) March 14, 2023

Had my wife close her eyes and describe rooms in her dream home.

In 3 minutes, I made these on Midjourney (h/t @nickfloats)

She was speechless.

She’s now a huge fan of AI and “gets it”

This is the way. pic.twitter.com/sjf3CQYtwa

— Kallaway (@kanekallaway) March 6, 2023

HOW DUOLINGO’S AI LEARNS WHAT YOU NEED TO LEARN — from spectrum.ieee.org by Klinton Bicknell, Claire Brust, and Burr Settles
The AI that powers the language-learning app today could disrupt education tomorrow

Excerpt:

It’s lunchtime when your phone pings you with a green owl who cheerily reminds you to “Keep Duo Happy!” It’s a nudge from Duolingo, the popular language-learning app, whose algorithms know you’re most likely to do your 5 minutes of Spanish practice at this time of day. The app chooses its notification words based on what has worked for you in the past and the specifics of your recent achievements, adding a dash of attention-catching novelty. When you open the app, the lesson that’s queued up is calibrated for your skill level, and it includes a review of some words and concepts you flubbed during your last session.

The AI systems we continue to refine are necessary to scale the learning experience beyond the more than 50 million active learners who currently complete about 1 billion exercises per day on the platform.

Although Duolingo is known as a language-learning app, the company’s ambitions go further. We recently launched apps covering childhood literacy and third-grade mathematics, and these expansions are just the beginning. We hope that anyone who wants help with academic learning will one day be able to turn to the friendly green owl in their pocket who hoots at them, “Ready for your daily lesson?”

Also relevant/see:

Duolingo turned to OpenAI’s GPT-4 to advance the product with two new features: Role Play, an AI conversation partner, and Explain my Answer, which breaks down the rules when you make a mistake, in a new subscription tier called Duolingo Max.

“We wanted AI-powered features that were deeply integrated into the app and leveraged the gamified aspect of Duolingo that our learners love,” says Bodge.

Also relevant/see:

The following is a quote from Donald Clark’s posting on LinkedIn.com today:

The whole idea of AI as a useful teacher is here. Honestly it’s astounding. They have provided a Socratic approach to an algebra problem that is totally on point. Most people learn in the absence of a teacher or lecturer. They need constant scaffolding, someone to help them move forward, with feedback. This changes our whole relationship with what we need to know, and how we get to know it. Its reasoning ability is also off the scale.

We now have human teachers, human learners but also AI teachers and AI that learns. It used to be a diad, it is now a tetrad – that is the basis of the new pedAIgogy.

Personalised, tutor-led learning, in any subject, anywhere, at any time for anyone. That has suddenly become real.

Also relevant/see:

Introducing Duolingo Max, a learning experience powered by GPT-4 — from blog.duolingo.com

Excerpts:

We believe that AI and education make a great duo, and we’ve leveraged AI to help us deliver highly-personalized language lessons, affordable and accessible English proficiency testing, and more. Our mission to make high-quality education available to everyone in the world is made possible by advanced AI technology.

Explain My Answer offers learners the chance to learn more about their response in a lesson (whether their answer was correct or incorrect!)
…
Roleplay allows learners to practice real-world conversation skills with world characters in the app.

Six Important Disciplines for Using AI in Learning & Development [Cavalier]

On 03/16/2023, in Artificial Intelligence/Agents/LLMs/and Related, business, content development, aggregation, repositories, corporate / business world, corporate universities / corporate training, experimentation, future, human-computer interaction (HCI), ideas, innovation, instructional design, intelligent systems, intelligent tutoring, learning ecosystem, Learning Experience Design, multimedia, training / L&D, workplace, by Daniel Christian

Six Important Disciplines for Using AI in Learning & Development — from linkedin.com by Josh Cavalier

Excerpts (emphasis DSC):

In practice, L&D professionals must responsibly use AI-based tools. AI outputs that introduce bias or falsehoods may adversely affect the learning process of individual associates as well as the company’s overall efficiency. A human must review all content before being implemented in a learning solution.
…
AI will revolutionize corporate learning and development because of the ability to analyze and process large amounts of data. These capabilities will enable learners to acquire knowledge and skills more efficiently while providing personalized learning pathways customized to their current skill levels and goals.

For example, AI-powered chatbots can provide immediate feedback to front-line sales looking for specific client solutions or trying to learn about new products. By fine-tuning models, AI can ingest vendor opportunities and dynamically priced products at scale, allowing a sales team to focus on data-driven solutions for their customers.

The creation of educational media will be automated through generative AI, including training videos, podcasts, and, eventually, eLearning courses. This new reality will level the playing field for instructional designers needing more media creation skills, ultimately allowing Learning & Development to create personalized, immersive learning experiences representing the entire learning journey.

Also relevant/see:

How artificial intelligence can support knowledge management in organizations — from realkm.com by Bruce Boyes

Examples of use cases:

Forecast sales probabilities
Discover organization inefficiencies by analyzing CRM records
Organize and summarize legal precedents relevant to a new case
Retrieve dispersed nuggets of information related to a troubleshooting situation
Facilitate feedback and peer review on communication systems (e.g., Slack)
Facilitate real-time smart sharing between marketing channels and sales pipelines
Find and apply question-answer pairs in online manuals to manage service knowledge
Provide more human-centered and accessible applications of knowledge through chatbots

ChatGPT as a teaching tool, not a cheating tool [Rose]

On 03/14/2023, in 21st century, 24x7x365 access, adjunct faculty, Artificial Intelligence/Agents/LLMs/and Related, colleges, digital learning, education technology, emerging technologies, experimentation, faculty, future, future of higher education, game-changing environment, higher education, human-computer interaction (HCI), ideas, innovation, instructional design, IT in HE, law schools, learning, learning agents, learning ecosystem, Learning Experience Design, Learning from the Living [Class] Room, liberal arts, pedagogy, personalized/customized learning, platforms, productivity / tips and tricks, professional development, teaching & learning, tools, UK, United States, universities, by Daniel Christian

ChatGPT as a teaching tool, not a cheating tool — from timeshighereducation.com by Jennifer Rose
How to use ChatGPT as a tool to spur students’ inner feedback and thus aid their learning and skills development

Excerpt:

Use ChatGPT to spur student’s inner feedback
One way that ChatGPT answers can be used in class is by asking students to compare what they have written with a ChatGPT answer. This draws on David Nicol’s work on making inner feedback explicit and using comparative judgement. His work demonstrates that in writing down answers to comparative questions students can produce high-quality feedback for themselves which is instant and actionable. Applying this to a ChatGPT answer, the following questions could be used:

Which is better, the ChatGPT response or yours? Why?
What two points can you learn from the ChatGPT response that will help you improve your work?
What can you add from your answer to improve the ChatGPT answer?
How could the assignment question set be improved to allow the student to demonstrate higher-order skills such as critical thinking?
How can you use what you have learned to stay ahead of AI and produce higher-quality work than ChatGPT?

Exploring generative AI and the implications for universities — from universityworldnews.com

Excerpt:

This is part of a weekly University World News special report series on ‘AI and higher education’. The focus is on how universities are engaging with ChatGPT and other generative artificial intelligence tools. The articles from academics and our journalists around the world are exploring developments and university work in AI that have implications for higher education institutions and systems, students and staff, and teaching, learning and research.

5 Ideas To Incorporate AI In Your eLearning Course — from elearningindustry.com by Christopher Pappas

Summary:

Artificial Intelligence is now taking the world of learning by storm. Here are 5 ways you can successfully incorporate AI in online learning.

Let’s say you’re training sales reps on handling different customer personalities. You can use this technology to diversify your branching scenarios so that trainees can also speak and not only type. This way, not only will the training become more realistic, but you’ll also be able to assess and work on additional elements, such as tone of voice, volume, speech tempo, etc.

These 20 jobs are the most “exposed” to AI, ChatGPT, researchers say + other AI-related items

On 03/11/2023, in Africa, Artificial Intelligence/Agents/LLMs/and Related, Asia, Australia, Canada, career development, change, Europe, global / globalization, India, society, UK, United States, workplace, by Daniel Christian

These 20 jobs are the most “exposed” to AI, ChatGPT, researchers say — from cbsnews-com.cdn.ampproject.org

Excerpt:

New research examining the effects of language modeling AI like ChatGPT on different occupations and industries finds that certain jobs, like telemarketers and teachers, are more “exposed” to the technology than others, such as psychologists and counselors.

Also relevant/see:

Must read: the 100 most cited AI papers in 2022 — from zeta-alpha.com by Sergi Castella i Sapé; with titles, citation counts, and affiliations.

How will Language Modelers like ChatGPT Affect Occupations and Industries? — from papers.ssrn.com by Edward W. Felten, Manav Raj, and Robert Seamans

An inside look at a 5th grade classroom using ChatGPT [abcnews.go.com]

On 03/11/2023, in Artificial Intelligence/Agents/LLMs/and Related, content development, aggregation, repositories, creativity, education, education technology, experimentation, future, human-computer interaction (HCI), ideas, intelligent systems, K-12 related, learning, learning ecosystem, teachers, tools, United States, vendors, by Daniel Christian

An inside look at a 5th grade classroom using ChatGPT — from abcnews.go.com
Kentucky teacher Donnie Piercey uses ChatGPT with his fifth frade students.

Excerpt:

“Like every other educator, I had that concern. Is this something that students are only going to use to cheat? So I started to think about like, ‘OK, what role is AI — artificial intelligence — going to play in the classroom?'” Piercey, the 2021 Kentucky Teacher of the Year, told “Good Morning America.” “And the more that I thought about it, I realized that there’s a lot more good that can come about through AI as opposed to negative things that can come in the classroom.”

The Librarian: Can we prompt ChatGPT to generate reliable references? — from drphilippahardman.substack.com by Dr. Philippa Hardman

Lessons Learned

Always assume that ChatGPT is wrong until you prove otherwise.
Validate everything (and require your students to validate everything too).
Google Scholar is a great tool for validating ChatGPT outputs rapidly.
The prompt works better when you provide a subject area, e.g. visual anthropology, and then a sub-topic, e.g. film making.
Ignore ChatGPT’s links – validate by searching for titles & authors, not URLs.
Use intentional repetition, e.g. of Google Scholar, to focus ChatGPT’s attention.
Be aware: ChatGPT’s outputs end at 2021. You need to fill in the blanks since then.

Did We Just Change Animation Forever? [corridordigital.com]

On 03/07/2023, in animation, Artificial Intelligence/Agents/LLMs/and Related, content development, aggregation, repositories, creativity, digital storytelling, experimentation, graphics, media/film, productivity / tips and tricks, skills, story, storytelling, by Daniel Christian

‘ChatGPT Already Outperforms a lot of Junior Lawyers’: An Interview With Richard Susskind — from law.com by Laura Beveridge
For the last 20 years, the U.K. author and academic has been predicting that technology will revolutionise the legal industry. With the buzz around generative AI, will his hypothesis now be proven true?

Excerpts:

For this generation of lawyers, their mission and legacy ought to be to build the systems that replace our old ways of working, he said. Moreover, Susskind identified new work for lawyers, such as legal process analyst or legal data scientist, emerging from technological advancement.

“These are the people who will be building the systems that will be solving people’s legal problems in the future.

“The question I ask is: imagine when the underpinning large language model is GPT 8.5.”

Blue J Legal co-founder Benjamin Alarie on how AI is powering a new generation of legal tech — from canadianlawyermag.com by Tim Wilbur

Excerpts:

We founded Blue J with the idea that we should be able to bring absolute clarity to the law everywhere and on demand. The name that we give to this idea is the legal singularity. I have a book with assistant professor Abdi Aidid called The Legal Singularity coming out soon on this idea.

The book paints the picture of where we think the law will go in the next several decades. Our intuition was not widely shared when we started the book and Blue J.

Since last November, though, many lawyers and journalists have been able to play with ChatGPT and other large language models. They suddenly understand what we have been excited about for the last eight years.

Neat Trick/Tip to Add To Your Bag! — from iltanet.org by Brian Balistreri

Excerpt:

If you need instant transcription of a Audio File, Word Online now allows you to upload a file, and it will transcribe, mark speaker changes, and provide time marks. You can use video files, just make sure they are small or office will kick you out.

Generative AI Is Coming For the Lawyers — from wired.com by Chris Stoken-Walker
Large law firms are using a tool made by OpenAI to research and write legal documents. What could go wrong?

Excerpts:

The rise of AI and its potential to disrupt the legal industry has been forecast multiple times before. But the rise of the latest wave of generative AI tools, with ChatGPT at its forefront, has those within the industry more convinced than ever.

“I think it is the beginning of a paradigm shift,” says Wakeling. “I think this technology is very suitable for the legal industry.”
…
The technology, which uses large datasets to learn to generate pictures or text that appear natural, could be a good fit for the legal industry, which relies heavily on standardized documents and precedents.

“Legal applications such as contract, conveyancing, or license generation are actually a relatively safe area in which to employ ChatGPT and its cousins,” says Lilian Edwards, professor of law, innovation, and society at Newcastle University. “Automated legal document generation has been a growth area for decades, even in rule-based tech days, because law firms can draw on large amounts of highly standardized templates and precedent banks to scaffold document generation, making the results far more predictable than with most free text outputs.”

But the problems with current generations of generative AI have already started to show.

You are not a parrot — from nymag.com by Elizabeth Weil and Emily M. Bender

Excerpts:

A handful of companies control what PricewaterhouseCoopers called a “$15.7 trillion game changer of an industry.” Those companies employ or finance the work of a huge chunk of the academics who understand how to make LLMs. This leaves few people with the expertise and authority to say, “Wait, why are these companies blurring the distinction between what is human and what’s a language model? Is this what we want?”
…
Bender knows she’s no match for a trillion-dollar game changer slouching to life. But she’s out there trying. Others are trying too. LLMs are tools made by specific people — people who stand to accumulate huge amounts of money and power, people enamored with the idea of the singularity. The project threatens to blow up what is human in a species sense. But it’s not about humility. It’s not about all of us. It’s not about becoming a humble creation among the world’s others. It’s about some of us — let’s be honest — becoming a superspecies. This is the darkness that awaits when we lose a firm boundary around the idea that humans, all of us, are equally worthy as is.

Introducing Q-Chat, the world’s first AI tutor built with OpenAI’s ChatGPT [Bayer]

On 03/06/2023, in 21st century, 24x7x365 access, Artificial Intelligence/Agents/LLMs/and Related, assessment, digital learning, education, education technology, emerging technologies, engagement / engaging students, experimentation, game-changing environment, homeschooling/homeschoolers, human-computer interaction (HCI), ideas, innovation, instructional design, intelligent systems, intelligent tutoring, interactivity, K-12 related, languages and translation, learner profiles, learning, learning agents, learning ecosystem, Learning Experience Design, Learning from the Living [Class] Room, learning preferences, more voice more choice more control, participation, personalized/customized learning, platforms, productivity / tips and tricks, story, storytelling, student-related, teachers, teaching & learning, tools, vendors, virtual tutoring, vision/possibilities, by Daniel Christian

Introducing Q-Chat, the world’s first AI tutor built with OpenAI’s ChatGPT — from quizlet.com by Lex Bayer

Excerpt:

Modeled on research demonstrating that the most effective form of learning is one-on-one tutoring1, Q-Chat offers students the experience of interacting with a personal AI tutor in an effective and conversational way. Whether they’re learning French vocabulary or Roman History, Q-Chat engages students with adaptive questions based on relevant study materials delivered through a fun chat experience. Pulling from Quizlet’s massive educational content library and using the question-based Socratic method to promote active learning, Q-Chat has the ability to test a student’s knowledge of educational content, ask in-depth questions to get at underlying concepts, test reading comprehension, help students learn a language and encourage students on healthy learning habits.

How ChatGPT is going to change the future of work and our approach to education — from livemint.com

From DSC:
I thought that the article made a good point when it asserted:

The pace of technological advancement is booming aggressively and conversations around ChatGPT snatching away jobs are becoming more and more frequent. The future of work is definitely going to change and that makes it clear that the approach toward education is also demanding a big shift.
…
A report from Dell suggests that 85% of jobs that will be around in 2030 do not exist yet. The fact becomes important as it showcases that the jobs are not going to vanish, they will just change and most of the jobs by 2030 will be new.

The Future of Human Agency — from pewresearch.org by Janna Anderson and Lee Rainie

Excerpt:

Thus the question: What is the future of human agency? Pew Research Center and Elon University’s Imagining the Internet Center asked experts to share their insights on this; 540 technology innovators, developers, business and policy leaders, researchers, academics and activists responded. Specifically, they were asked:

By 2035, will smart machines, bots and systems powered by artificial intelligence be designed to allow humans to easily be in control of most tech-aided decision-making that is relevant to their lives?

The results of this nonscientific canvassing:

- 56% of these experts agreed with the statement that by 2035 smart machines, bots and systems will not be designed to allow humans to easily be in control of most tech-aided decision-making.
- 44% said they agreed with the statement that by 2035 smart machines, bots and systems will be designed to allow humans to easily be in control of most tech-aided decision-making.

What are the things humans really want agency over? When will they be comfortable turning to AI to help them make decisions? And under what circumstances will they be willing to outsource decisions altogether to digital systems?

The next big threat to AI might already be lurking on the web — from zdnet.com by Danny Palmer; via Sam DeBrule
Artificial intelligence experts warn attacks against datasets used to train machine-learning tools are worryingly cheap and could have major consequences.

Excerpts:

Data poisoning occurs when attackers tamper with the training data used to create deep-learning models. This action means it’s possible to affect the decisions that the AI makes in a way that is hard to track.

By secretly altering the source information used to train machine-learning algorithms, data-poisoning attacks have the potential to be extremely powerful because the AI will be learning from incorrect data and could make ‘wrong’ decisions that have significant consequences.

Why AI Won’t Cause Unemployment — from pmarca.substack.com by Marc Andreessen

Excerpt:

Normally I would make the standard arguments against technologically-driven unemployment — see good summaries by Henry Hazlitt (chapter 7) and Frédéric Bastiat (his metaphor directly relevant to AI). And I will come back and make those arguments soon. But I don’t even think the standand arguments are needed, since another problem will block the progress of AI across most of the economy first.

Which is: AI is already illegal for most of the economy, and will be for virtually all of the economy.

How do I know that? Because technology is already illegal in most of the economy, and that is becoming steadily more true over time.

How do I know that? Because:

From DSC:
And for me, it boils down to an inconvenient truth: What’s the state of our hearts and minds?

AI, ChatGPT, Large Language Models (LLMs), and the like are tools. How we use such tools varies upon what’s going on in our hearts and minds. A fork can be used to eat food. It can also be used as a weapon. I don’t mean to be so blunt, but I can’t think of another way to say it right now.

Do we care about one another…really?
Has capitalism gone astray?
Have our hearts, our thinking, and/or our mindsets gone astray?
Do the products we create help or hurt others? It seems like too many times our perspective is, “We will sell whatever they will buy, regardless of its impact on others — as long as it makes us money and gives us the standard of living that we want.” Perhaps we could poll some former executives from Philip Morris on this topic.
Or we will develop this new technology because we can develop this new technology. Who gives a rat’s tail about the ramifications of it?

Categories

Tags