We have had another wet week. We now have the exact opposite of the weather that we experienced last summer. This is typical of our weather, when we need rain, it stays dry and when we could do with a bit of a break in the rain it just won’t stop.
It’s far too wet now to do any gardening, so we have had to look towards alternative occupations.
Whenever there has been a break in the rain, we have taken the opportunity to go walking with Ruby, although our walks have been local.
Both Martyn and I have been exploring Ai. I have been exploring two apps. The first that I tried was ChatGPT. I used it to get some recipes, calculating a diet for a friends overweight beagle and researching a few other things. The information produced and the conclusions made would have taken me ages if I’d had to search for and assimilate all the information provided.
One recipe I asked for was for some biscuits. I asked for a recipe that used dried cherries and blueberries. The biscuits didn’t turn out very biscuit-like, although the taste was OK, the shape was more like a small cake.
I uploaded the photo to ChatGTP to ask why the recipe hadn’t worked as expected, and was given several possible reasons. ChatGPT decided that the brand of flour had made the difference and so it edited the recipe specifically for the brand of flour that I was using. I haven’t tried that recipe yet.
We always make our own Christmas cards, so I decided to ask ChatGPT to help with the design. I used a photo of Ruby and asked ChatGPT to put her in various Christmassy backgrounds and outfits
ChatGPT was much slower at producing images and only allowed a limited amount before I was told that I couldn’t do any more until the next day. This is the free version so I guess that this is to encourage you to subscribe. There was also some distortion on Ruby’s image. It lost some of her character.
I had read that another app – Gemini – produced better images so I decided to try that. I took a photo that I thought would make a good starting point.
I loaded this into Gemini and asked for Ruby’s harness to be removed. Then I asked Gemini to provide a range of backgrounds based on a description that I gave.
I asked for Ruby to be dressed in various Christmas outfits.
I asked for text to be added, and all along the way, I could tweak various parts of the image and change the backgrounds, all of which were generated quickly.
As well as changing backgrounds and outfits I could also add items to the image like a poodle styled snowman, or move things about
I much preferred Gemini when working with images and ended up with dozens to choose from. Eventually was told to take a break as I had created too many images and I had to wait until the next day to continue. I didn’t choose any of the above images for our card. I’ll reveal that in a later post.
When it came to printing out our cards we hit a problem. After printing one side of the cards the printer refused to pick up the card. For ages we fiddled around with it to no avail and so I decided to consult ChatGPT. In a matter of seconds, it came up with several possible causes of the problem out of which a worn roller was the most likely problem.
ChatGPT asked for the printer model and then told us how to get around the problem by using a manual paper feed at the back of the printer that we didn’t even know existed. We were given instructions as to how to access it. Bingo, we had print off.
My conclusion with regards to the two Ai apps is that Gemini is better for image creation and ChatGPT for accessing information.
When I wasn’t designing Christmas cards or quizzing ChatGPT, I spent time reading a new Kindle book.
A while ago we decided to subscribe to Amazon Prime and one benefit is access to a free Kindle library. The library holds a good selection of titles. The books change, so there is always a good selection. I can ‘borrow’ up to’ ten titles at once for an indefinite length of time. After reading the book it is then returned to the library. Prime becomes familiar with the types of books that I read and makes suggestions based on this. Previously, I bought ten titles at a time and my library of titles bought is extensive. I rarely read the same title twice so the idea of borrowing titles suits me. I can keep my Prime library stocked up so that I always have something to read.
Finally, one indoor game that we play with Ruby, is find the kibble.
If you want to watch Ruby playing this game, a video is posted here.
That’s all for this week, so as always, until next time stay safe, stay well, stay warm and stay dry.
Awww, I like all those photos of Ruby, AI is working very well for you. Perhaps your biscuit recipe is based on what a biscuit is like in America, more of a scone.
It was a cookie recipe and told me the problem was with the flour that I use, Jo. Apparently Homepride needs smore liquid – who knew? It’s given me an edited version that should work with my flour – we will see.
You’ve mastered working with AI very well, Sue. Looking at the first picture, I thought it was actually snowing!
Did you thunk we dressed Ruby uo too, Nadezda? I don’t thinks she would tolerate a woolly hat