Month: May 2023

Losing Value (What will AI reduce in value)

When I was young, I learned how to take photographs with a 35mm SLR camera, take the film out of camera, process it and create prints. I was fortunate my dad was interested in photography and built a lab in our basement. The cost of a picture was not cheap, the film, the processing, the time it took to do it all. Even for those who didn’t process their own pictures you needed to take them to a lab to get them turned into prints. Now there are polaroid cameras which were instant, but the costs were high. Of course, it got cheaper over time, and some fast processing but the costs of each picture and doing it was a challenge. 

In 1975 Kodak created the first camera and what followed was a slow reduction of costs in pictures. In 1989, the first commercially available digital camera to be released (Fuji) The cameras were expensive, the resolution started out very poor, but technology would advance. 

In 1997 a cell phone was released with a camera, and about the same time online photo sharing sites started going online. Ofoto, Shutterfly and DotPhoto allowed people to upload photos and share with others. Instead of printing pictures out and putting them in albums users could upload and share them. Camera companies had these smaller (pocket size) digital cameras trying to make it easier to get digital pictures, as well as they had ways to upload them. This was still time consuming and took some technical skill. 

In 2006 Facebook was available to all users (over 13), Twitter came out and in 2010 Instagram was released. Even though social networks were around before this, it wasn’t till these three figured out how to get mobile photos that were about to change.  

Well thanks for the history lesson, Larry, but I could have gotten that all from Wikipedia, in fact it might be more correct. Absolutely and you should check my work, but believing everything you see on the internet is another post. Let us connect the dots, people have phone which now has a good enough camera, (yeah, I didn’t put the iPhone in there, but you know when that came out) companies have made it easier to get those photos online (social media etc.) so the result was people started taking more pictures with very little friction (cost, knowledge etc.)  

The cost of taking pictures from the old days of film cameras you had to buy, get film, get it developed etc. had diminished to zero. Photography had crossed the chasm from just a few to mass adoption. Before you only took pictures at moments (how many birthday pictures do you have from the 60-70s) now you take pictures at almost every meal. If fact, my camera roll has so many pictures there is no way I can ever look at them all. The days of sitting on a couch and going thru a photo album pretty much is over.  

What does his have to do with AI? Over the last few months as you play with Chat GPT and other tools and some people look at it for what jobs it will take away, others look at the damage it can possibly do and lastly those who want to figure out how to get the best out of them. The one point I think about is what is this going to make so cheap its worthless. To me over time many things have hit this. 

  • Photography (as mentioned) 
  • SPAM (well mail) when you can email something at no cost, vs price of a stamp 
  • Travel, Food etc, Guides. Remember buying books on places to travel 
  • Navigation, remember going to AAA and getting a trip-tick? 
  • Chilton’s Car repair guides. A simple YouTube search has made this really free 
  • Wikipedia killed Encyclopedias  
  • No one knows what cliff / monarch notes are 
  • Why don’t you add your own… 

What is General AI going to make worth less. First, it is not 100% correct. Pre-Chat-GPT a google search required you to read some links and decide what a good answer (or confirmation bias) to your question is. I wrote about AI being the next GPS and proof is this article about a lawyer who not only used ChatGPT for his case, but he only check the sources by asking ChatGPT if it was lying.  

Though not sure anyone saw the photo becoming worthless now we don’t need to think if it is worth taking a picture as the cost is zero. With AI, we can only wait to see what makes it so friction free (writing stories, creating images etc.) that there is little value. I see the productivity gains leveraging it (images for this blog, coding etc.) and be watching to see.  

I love and hate what happened to photography. But looking at the list of things that are free/frictionless and most of this made my life better so now I will continue to experiment and find the next thing that will be friction free. 

This opinion is mine, and mine only, my current or former employers have nothing to do with it. I do not write for any financial gain, I do not take advertising and any product company listed was not done for payment. But if you do like what I write you can donate to the charity I support (with my wife who passed away in 2017) Morgan Stanley’s Children’s Hospital or donate to your favorite charity. I pay to host my site out of my own pocket, my intention is to keep it free.  I do read all feedback, I mostly wont post any of them

This Blog is a labor of love, and was originally going to be a book.  With the advent of being able to publish yourself on the web I chose this path.  I will write many of these and not worry too much about grammar or spelling (I will try to come back later and fix it) but focus on content.  I apologize in advance for my ADD as often topics may flip.  I hope one day to turn this into a book and or a podcast, but for now it will remain a blog.  

Geoffrey Moore was right about the Chasm

One of my favorite books’ pre-dates the wonderful Dot-Com explosion and bubble, and one of my favorite authors has a book that I find places to apply the ideas in that one book. I wrote previously about Cult is the root of culture, but I really missed a single point.  

Changing culture is done in large corporations by having a marketing campaign and creating presentations about change. Maybe finding one or two senior managers who can act as ‘change agents.’ Does this work? (Maybe you should answer that question not me.) 

What Geoffrey Moore talks about in his book has to do with innovation, but I believe anytime there is a needed change the model works well. His book describes the diffusion of innovations which goes back to 1962 by Everett Rogers.  

Yep it’s a bell curve… And as much as people gripe and moan it works. Everyone thinks they are good drivers, but realistically most of us are somewhere in the average. But this curve has two far ends, one with innovators (Steve Jobs, Elon Musk etc.) as the top say 2-3% and the next 10-15% as early adopters (those standing in line to get some new tech.). The rest (the 68% of the people in the above average/below average) and the people trailing are going to wait till whatever change has crossed the chasm.  

That first 12-18% who are willing to change are the people that you need to focus on. If you are trying to change culture, change a process, change anything, these are the people who are willing to take the risk. You need to find those people, and work only with only those people. Treat it like new tech that is ‘invite only’ an exclusive club. If you get that first generation to change the notion of that exclusivity adds to the FOMO (Fear of Missing Out.). The middle 68% will see the success of that top group and will want in.  

Finding those people depends on what you are trying to change, but obviously anyone who pushes back is not one of those people. And as said make it invite only, and even better make it a challenge to get into the group (so they want to be there) will help drive the desire to make the change. This is not easy, and others will say change is not easy period. But using the diffusion of innovation will give you a chance to succeed, more than any marketing campaign for people who say ‘I will wait to see if anyone else gets it to work’. 

So if you want change, culture, style of work or other find the visionaries have them help build the change, find the early adopters get them on board and the rest will eventually follow.

This opinion is mine, and mine only, my current or former employers have nothing to do with it. I do not write for any financial gain, I do not take advertising and any product company listed was not done for payment. But if you do like what I write you can donate to the charity I support (with my wife who passed away in 2017) Morgan Stanley’s Children’s Hospital or donate to your favorite charity. I pay to host my site out of my own pocket, my intention is to keep it free.  I do read all feedback, I mostly wont post any of them

This Blog is a labor of love, and was originally going to be a book.  With the advent of being able to publish yourself on the web I chose this path.  I will write many of these and not worry too much about grammar or spelling (I will try to come back later and fix it) but focus on content.  I apologize in advance for my ADD as often topics may flip.  I hope one day to turn this into a book and or a podcast, but for now it will remain a blog.   New Images are created using AI with license to use.

Do you need a fall back plan to chase your dreams?

When I was in college, I did take a sideways direction as I realized I was not talented enough to be a professional musician but had the brains to manage bands. I took a course given in NYC about it, and there was so much information it blew my mind. The presenter (I apologize this was 87 I do not remember his name) had some interesting advice, sometimes it was contradictory. 

In the section about working with a band and the band needing to ‘self-fund themselves’ he told the story about one band that took one of the members grandmother’s fudge recipes and started selling it to fund the band. The idea was to give the band enough time to practice full time do chase their dreams. The band did not make it but ended up opening a fudge store as a full-time gig (I had to chase that down later.) Thank God I did not quit my day job…. Well, like George. 

In recent days conversations with my daughters touched on do they have a backup plan if they do not get into Med or PA school. In this case I was the pragmatic father and wanting to make sure they had a plan b. This also could be related to my up brining where my parents wanted to make sure I had a stable job etc. I could chase dreams, but my mom always said have a plan b, c, d….  

Reading what I had written previously, I noticed I wrote about Failing at something you don’t like as a reason to chase your dreams. In a research project the notion of having a fallback plan has some drawbacks. So, the downsides? 

  • Not going ‘All In’ on your dream reduces the effort you can put in 
  • Wasting resources on backup plan that could be used on your primary plan 
  • Apparently just thinking about a backup plan reduces your chance for success 
  • Falling into your backup plan due to the sunk cost fallacy 

Now the studies talk about how people just think about goals, or in some cases have small ‘monetary’ incentives. Looking back at my desire to be in the music business, I did focus a lot on a Plan B (like my mom told me.) Would I have made it with Plan A? I really do not know, but my focus on Plan B took me far. 

Now back to what I told my kids. The first make your mistakes in your 20s (well now) It is much easier to find an alternate path if you make the mistake now. Trust that you have the resources (parents/friends) to help you out, and your own abilities to find your Plan B. So, you will not hit FIRE at 40, finding happiness and your place in this world may be more important than retiring early.  

And now to think about how to do this at work. Many times, I have been asked if there is a Plan B, and now my response should be if we focus on it, it may turn into Plan A. 

This opinion is mine, and mine only, my current or former employers have nothing to do with it. I do not write for any financial gain, I do not take advertising and any product company listed was not done for payment. But if you do like what I write you can donate to the charity I support (with my wife who passed away in 2017) Morgan Stanley’s Children’s Hospital or donate to your favorite charity. I pay to host my site out of my own pocket, my intention is to keep it free. I do read all feedback, I mostly wont post any of them

This Blog is a labor of love, and was originally going to be a book.  With the advent of being able to publish yourself on the web I chose this path.  I will write many of these and not worry too much about grammar or spelling (I will try to come back later and fix it) but focus on content.  I apologize in advance for my ADD as often topics may flip.  I hope one day to turn this into a book and or a podcast, but for now it will remain a blog.

Agile and the New Pair Programming Paradigm with GPT-AI

Yes, I know for the last hundred plus posts it has been about getting better. But over the last four months I got into playing around with AI. Well, I hate to call it AI, GPT is a better name (Generative Pre-trained Transformer.) The speed of announcements and actual tools are coming so fast it boggles the mind. 

First if you do not know the history of Agile, nor read the Extreme Programming Explained book, please google that, and get some back history. One key concept in EPE is the notion of pair programming, where two developers sit together to work on a single piece of code. There is a lot of good to this, one developer can catch the other typos, opine on unit tests, and in many cases two brains are better than one. This part of Agile is what I am going to say should change based on the tools I have played with. Note I have no investment in any of these tools and will focus on the free ones for this writing. 

So, the outcome of pair programming is better code quality, faster problem solving, knowledge sharing (no single point of failure) and reduced on boarding time. A developer can be productive from day one via pair programming. 

Now pair programming does have some negatives. Finding two developers that work well together is not easy, even finding a programmer that ‘wants a partner’ is hard enough to find. Add scheduling, developers working remote (Covid) and some companies having global presence. There are even more reasons why pair programming did not take off,  

Now the curve ball, along comes Chat-GPT. And even a non-programmer can send a simple request and it will spill out code. No developer needed! Now, of course, the problem is making sure that code works, does what it says it does, has all the non-functional requirements etc. With us geeks (programmers) we have got our hands on some new tools. Codeium, Code Whisperer, Git Hub CoPilot etc. show up. And presto we now have tools to help us code. Of course, they suffer from some bad reasoning that GPT tools have. 

Some of the tools (Codeium) have a ‘chat function’ where you can ask it to write code. And let’s say now you have the business user sitting with you and giving you, some requirements (user stories) and you translate it to the correct prompts. The tool writes some code, as a developer you can enrich the code. The next step is to ask the tool to write test cases, and presto they are done.  

Of course, a developer can choose to write code and allow the tools to give suggestions to speed up the coding and correct mistakes. Again, no need for the second developer to do that. Let the GPT do these things for you. The tool is available all the time, does not have to worry about sick days, regions etc. These tools are also faster than the second programmer and can make the changes in real time where the second programmer will have to say, ‘correct your mistake.’ One last key feature is it can enforce standards (naming, design etc.)  

All of this at much less cost than the second programmer. For the same costs (about) the second programmer can be working on something else with his/her own coding tools.  

Now Agile may change, not just pair programming, but think of moving to Kanban where now the product owner (or a delegate) sits with the developer and just talks through ideas while the coder developers in real time. The possibility of productivity gain is unreal.  

There are some downsides as these tools do not have a record of accomplishment yet, setting it up for your code only, or sandboxing is coming to prevent your ideas and code being ingested in a public GPT must be perfected. And as my joke said, sometimes the idea can be completely wrong (hence the need for the first programmer.) 

I see the positives and wonder how soon it may become reality. There is this push of AI dangers, the possible loss of jobs etc. but I prefer to figure out how to use it to get better. 

This opinion is mine, and mine only, my current or former employers have nothing to do with it. I do not write for any financial gain, I do not take advertising and any product company listed was not done for payment. But if you do like what I write you can donate to the charity I support (with my wife who passed away in 2017) Morgan Stanley’s Children’s Hospital or donate to your favorite charity. I pay to host my site out of my own pocket, my intention is to keep it free. I do read all feedback, I mostly wont post any of them

This Blog is a labor of love, and was originally going to be a book.  With the advent of being able to publish yourself on the web I chose this path.  I will write many of these and not worry too much about grammar or spelling (I will try to come back later and fix it) but focus on content.  I apologize in advance for my ADD as often topics may flip.  I hope one day to turn this into a book and or a podcast, but for now it will remain a blog.

© 2024 LrAu

Theme by Anders NorenUp ↑