I thought I might do something different today in my post.
Perplexity has become an essential tool in my paired programming toolbox over the last few weeks. The other programming tool being GitHub‘s Copilot.
I haven’t used AI to generate many images. Writing prompts to create Python scripts differs from scripts to generate images.
In my opinion, image creation requires more details up front than computer code. If that is incorrect, would you share your prompt iterations in the comment below by using a link similar to this one?
– https://lnkd.in/gctyrhY8 –
I was amused as Microsoft Copilot, using Designer powered by DALL-E 3 just stopped creating images after the first attempt. The evidence is shown here, at this link.
– https://lnkd.in/gQ6qN_BM –
In the productdevelopment space, there is a concept around “fit for purpose and fit for use”.
Generative AI has a purpose that it fits. Analyze the statistical sequence of words and calculate a probable response based on the word values.
A text response is fairly direct and can be steered in a desired direction with relative ease. Image generation, not so much.
In its usefulness for generating images, generative AI is advancing. But, it’s not quite where it could be just yet.
For paired programming focused on solving a generic problem like dataset comparison, AI does quite well at generating useful text-based responses.
Looking at generative AI through the product development lens, it might be said that it is fit for one or maybe two purposes. It also might be said it is fit for one or perhaps two uses.
Generative AI cannot solve all problems, like the experience with image generation. It can remove repetitive, tedious work that humans do not perform well that is better suited for machines.
Consider how your business or company might use AI to compress labor this week.
The image below was created using ChatGPT 3.5 and DALL-E on the OpenAI platform. Making me into a cartoon did not work out very well.
As a parent, a goal I aspired to was achieved earlier this week.
Our daughters are visiting their relatives in Croatia this summer. It’s a milestone since they travelled together with a friend and without my wife and me.
Mission accomplished, being able to speak with them from the comfort of our Dallas-Fort Worth home while they enjoyed desert with their grandmother at a local pastry shop in Zagreb.
As a product coach, I consider how we can scale capability in a given team. Product development should enable teams to evolve to a point where they are sufficient and self-reliant in their abilities and skills.
When I am able to step away from a team and cheer their success from the sidelines, I consider my mission completed.
Like with our children, coming alongside for a defined time period to achieve desire outcomes should be what a coach desires for their team.
I’m proud that our children have become capable, emerging adults. I’m equally proud when a team can confident say they no longer need me as their full-time coach.
How do you define success for your product teams and their coaches?
A productive morning moved on to be a productive day.
I won’t bore you with the details of my workday in this post. Instead, I’ll highlight the early morning start with the monthly Success North Dallas event.
The panel on AI was outstanding with Lori Nugent moderating, my friend and colleague Yeshwant Muthusamy, Ph.D. contributing alongside Larry Roberts and Chris Mathew their insights.
It’s wonderful to see our community having a conversation around the productive and safe use of artificial intelligence.
As we evolve product development, AI will become more integrated into what we build.
How we use it and when it makes sense to embed, those are critical questions we need to discuss today.
My closing question is, how might we partner with you to solve your painful technology problems using proven approaches?
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26
