Trying to have witty repartee with AI can feel like doing card tricks for a dog. I was delighted Copilot got the joke when I followed up a question about converting kph to mph with a question about furlongs per fortnight.
AI laughed at my joke
Snoqualmie Falls
My 2 question test for talking about Shakespeare
I’ll talk about Shakespeare with anyone willing to join me. The first time I tried Chat GPT, looking at the blank box on the screen with the intention of taking the thing for spin but not actually having anything in mind to task it with, I think I must have wanted to test whether it had any sense of irony. I threw some Shakespeare at it. “Should I ever a borrower or lender be?” And then it came back with a strong attempt to give me financial guidance for the 21st century. It didn’t reference Polonius passing on the advice to Hamlet. So my follow up was, “But isn’t Polonius an ass?” It came back with a strong “No” and more or less explained that his advice was wisdom that had been recognized for centuries. I moved on to other topics.
After about three and half years, I figured things had progressed enough that it was time to try again. And now I have more choices. I’m settling into a new personal computer and trying out AI tools, so I fired up Chat GPT, Claude, and Gemini. As before each tried to provide a framework of financial guidance for our contemporary world, but this time each did start out by referencing Polonious’ line. Responses to the follow up question, “But isn’t Polonious an ass?” did a good job explaining how the audience should read Polonious, and they mostly used the same language. Claude, was my favorite with it’s gleeful exclamation mark, “Yes, absolutely!” Gemini used boldface and hanging quotes “Most scholars and audiences would agree: yes, Polonius is absolutely an ‘ass.’” Chat GPT brought out all the punctuation and used italics “Short answer: yeah…he kind of is—but in a very intentional way.” I enjoyed Gemini describing the speech as “a collection of sententious platitudes” which sounds pretty damning. I also enjoyed Claude describing the “endless string of maxims” and Polonius “firing them off like fortune cookies.” Each had a good basic understanding of dramatic irony, and Claude even made a try at breaking the fourth wall: “the joke may be on everyone who quotes ‘Neither a borrower nor a lender be’ as genuine sage advice.”
Claude also put forward that Ian Holm had played Polonius when he was with the Royal Shakespeare company, but added “I’d also gently flag that I should have been more careful earlier — my claim about Ian Holm playing Polonius specifically was stated with more confidence than I can actually back up.” Some of the best arguments I’ve heard about how to interpret Shakespeare have been backed up with had nothing more than the inflection with which they were delivered, so I’m not going to hold Claude’s initial claim against it. I looked it up and now realize I need to watch the 1990 film where Mel Gibson played Hamlet and Ian Holm played Polonius. I bet Holm is fun. When he was with the RSC he did play two of the greatest asses in all of Shakespeare–Malvolio and Troilus. I’m sure I’ll get around to writing about both Twelfth Night and Troilus and Cressida sooner or later. And as the android in Alien, Holm was wonderfully ironic about being human.
Zani
Dragon Eye Pine
Google Cloud Vision, Art, and Odd Photos
To quickly give someone a look at the state of computer vision, there is a fun demo on the Google Cloud Vision site. You just drag a photo onto it, and you can learn what the AI sees along a few key vectors. After seeing it easily recognize and classify my photos of poodles and cherry trees, I started playing with photos I’ve taken that are actually tricky for humans to parse. Playing that type of game helps define the limits of what’s possible today, and things that happen at the boundaries suggest very interesting possibilities for tomorrow. I found that Cloud Vision today can touch on the subjective experiences of viewing art, and also do a good job sorting out things that are merely odd as opposed to things truly expressing artistic intent.
In this photo I laid a Christmas wreath on top of a pizza box that had an illustration of Sasquatch wearing a Seattle Sonics headband.
Cloud Vision nails the Christmas decoration aspect of the photo, but I was delighted to see that it indicated that there was about a 50% chance that this was some kind of art.
I’ll admit that it’s not very good art, but when I put together my little tableau, I was trying to capture something about the quirkiness of Seattle with an artifact I could use to communicate to people not from here. Personally, I’d set the odds of it being granted the status of “art” at much less than 50% so I felt Cloud Vision’s assessment was generous.
After that discovery through experimentation, I figured I should establish a control before doing more tests. Below is a photo from the Microsoft art collection. It’s a graffiti covered piece of the Berlin Wall on display at the Microsoft Conference Center in Redmond.
It’s been about 40 years since anyone has questioned whether or not graffiti art is worthy of gallery status—a sensibility Cloud Vision is certainly aware of based on how it labeled this piece.
Below is a photo I took just because I was thinking, “I sure have some odd stuff in my yard.” The metal thing is a watering trough for livestock that is destined to be a planter. The black things sticking up out of it are foam gutter inserts that keep leaves from blocking your gutters but let water flow through.
As a thing, it could be a Philistine’s parody of modern art (it bears a slight resemblance to Typewriter Eraser, Scale X by Claes Oldenburg) but not even the rubes would mistake it for actual art. Although I was a little surprised Cloud Vision couldn’t identify a watering trough, it did offer some pretty valid labels.
Mostly, I was heartened that it didn’t try to slap the label “art” on it.
The Assassination of Hope Fails Again
The image that used to be the header for this blog is a painting by Andy Kehoe entitled The Assassination of Hope Fails Again. It’s an allegorical pun. The deer-looking animal is “Hope.” You can’t kill him, because he springs eternal. <rimshot>
I use it here because I am inherently an optimist, and I enjoy the humor of the painting. I also like that it reminds me of eastern Oregon, one of my favorite places.
But, I think I mostly use it because its colors complement the gray and white style of this WordPress theme. Likewise, the way it connotes the old west works well with the way this theme uses slab fonts.
And of course I use it because I love the painting. It hangs on my wall at home.
Earlier blogging
Most of my blogging for the past few years has been published to Microsoft sites, and this is actually the third time I’ve spun up a personal blog.
Microsoft posts
Below are a few of my highlights from the Microsoft Learning, Machine Learning and Azure blogs. I have to give a big shout out to my collaborator and editor Shannon Wright for these as well. Her contributions were essential.
AI for Education: Individualized Code Feedback for Thousands of Students This post describes the code grader the team created in collaboration with Microsoft Research so that we could provide useful evaluations for thousands of student projects in our Introduction to C# course. My favorite part was when the light went on for me in regard to creating AI that understands the semantic meaning of a program rather than merely the syntax. It felt like the payoff for all my time spent studying Wittgenstein and Benjamin back in grad school.
Serious about cloud security? Check out this new training on Azure Security Center This one is a good representative example of how we communicated the real world usefulness of the training we made available. There is no topic where it’s more important to establish yourself as a trusted resource than security, and a key way to do that is to promise and deliver on walking people through examples of exactly how to use the tools in the real world.
Microsoft Cognitive Services: “The real magic is in the API.” –Scott J. Peterson This one covers our courses for getting hands on with Microsoft’s Cognitive Services so that you can, among other things, quickly create an app that can see and index the content of a video or photo or translate languages. The fact that computers can do these tasks does feel almost uncanny, but like Scott says, what’s really magical is that the power to use them is so accessible.
Get your team certified on cloud-critical identity and access management skills The second of a series of four posts focused on the value of Microsoft’s professional certifications from a business decision maker’s point of view, I like the way this one charts the course from identity as a problem to identity as a solution. Our intention was to provide a CEO with enough conversational knowledge to have a meaningful discussion with his IT team about their roadmap.
Personal blogs
My first blog. My first blog was on Typepad, because I had recently seen Anil Dash speak at a conference and I was blown away by his presentation. (To this day he continues to impress me, and I still pay close attention to his work.) There aren’t any posts that I’d call out as highlights, but I was happy to see it’s still live even though I haven’t touched it in years.
N – 1 blog. Like this blog, my second was a WordPress site. The posts there were timely, and several were written as part of ongoing conversations where I had more to say than could fit in email, but out of context and out of time, they don’t really bear up to reading today.
Super Fan Facebook Graph, is the exception. I took a course on mapping social networks just so I could get a picture of what my personal social graph looked like after spending a few years deeply embedded in the culture around the reality show Big Brother. The former contestants had huge numbers of connections to each other and to fans; likewise, the biggest fans also had massive numbers of friends and followers. The comparison to my organic groups of “people I play bridge with” and “people I went to school with” still boggles. I refer people to this post often.