Newspapers in the AI Era

Today, I saw an article on Linked in about how the Encyclopedia Brittanica is now positioning itself as an AI research company (https://www.linkedin.com/feed/update/urn:li:activity:7275863301777997824/).

This is a reminder that, with the AI Era, any company with access to data has a valuable asset. This makes me think of newspapers.

Major newspapers such as the Chicago Tribune, have over 100 years of articles, information, reviews, etc. I always thought that one thing that can save them is to partner with other media to present "infotainment", such as a streaming series on organized crime during Prohibition, etc.

Now, I think they can also become AI research tools. They can create interfaces that let people type in a prompt and it can present research from its archives, with the reference line generated. Perhaps they can set the paywall so that, someone clicking on the reference inserted in an paper than allows them to read only that article.

It would allow students and researchers to be able to expand the first-person resources they can use.

© 2024 Praveen Puri

Why It's Important to Separate Data from Business Logic

I recently advised a client with a critical data acquisition project.  It reminded me yet again why it's so important to abstract parts of a system as black boxes with interfaces.

In their case, they get data from many different sources, in different formats (such as web APIs, CSV files, pdfs, etc), then process them with business logic.

They recently had one critical source who suddenly, with no warning, changed the format of the information.  Instead of letting them rewrite the business logic, which would then require extensive QA testing, I showed them how to call an existing library function which converted the data back to the old format.  Once they did this, they were able to re-use the same business logic without modification, thus saving a lot of retesting.

© 2024 Praveen Puri

Technology Platform "Chicken and Egg Problem"

If you come out with a new hardware or software platform (like a new chip or new operating system), customers will not want to buy it unless it has a lot of useful apps.

App developers, however, won't write for your platform unless they see values, in terms of many customers.

Almost 25 years ago, Joel Spolsky wrote about this problem (https://www.joelonsoftware.com/2000/05/24/strategy-letter-ii-chicken-and-egg-problems/), and he feels that the solution to consider is backwards compatibility.

In other words, you spend a lot of time figuring out how your system will run apps from the current popular platforms.

For example, Microsoft DOS got established on IBM PCs because it offered backwards compatibility to CP/M, the operating system in use at the time, which had Word Star, a popular word processing application.

Also, competitors to Intel's x86 chips did not make any progress until they made their new chips capable of emulating x86 chips.

Spolsky also mentions the case of early, online bill payment services. The ones that succeeded offered "backwards compatibility" for merchants that weren't directly enrolled with them. They simply provided a mailing address where customers could have their bills forwarded, where they were manually scanned and uploaded.  It was tedious but, once they got a critical mass of customers, they could go back to the merchants and show them how much they can save by not having to mail paper bills.

© 2024 Praveen Puri

College Football is Now a Professional League

Now, college football has done what the USFL, XFL, etc. couldn't accomplish. It's now a professional, alternative league to the NFL. 

That is why Bill Belichick feels comfortable coaching in NC: it's much more like the NFL now.

© 2024 Praveen Puri

Consulting Haiku

Consulting goes with
the ancient art of haiku:
two sides, yin and yang

© 2024 Praveen Puri

Myth: People Don't Like Change

It's a myth that people don't like change.  If they didn't, they would never go to college, get married, or have kids.  What people don't like is to have change thrust on them, where they don't see the value in it.

If you want to change a workplace,  the workers need to help drive it.  They need to understand the change, why it's needed, and how it helps them and the company.  They need to help shape the details.

This means that you don't just spring a new computer system on them one day and provide them with a couple of hours of training.  This means, as you develop the system, you also develop simulators, and you allow groups of users to try them out (even at the beginning of the project, where functionality is minimum).  Get their feedback, and incorporate that into the design.  There are no excuses to not iterate, especially if you consider yourself "agile."

Finally, when the system is ready, you don't just tell the users to do everything with the new system, especially if they are under a time crunch.  Have both systems side by side, and start by letting them work a few, simple tasks through the new system.  Let them naturally shift over their workload to the new system as they get comfortable.

These are all covered by my Strategic Simplicity® Framework, CLOUD:

Change simplicity

Language simplicity

Operational simplicity

User simplicity

Decision simplicity 

© 2024 Praveen Puri

The Real Twilight Zone

The real Twilight Zone is when you wake up and see that there is less than 15 minutes before your alarm will go off.  Do you try to sleep or get up?

© 2024 Praveen Puri

Monitoring Independent Key Indicators

When my support team was monitoring complex financial applications that spanned multiple servers, processes, and software teams, we put in place dual monitoring. First, we had a page that was updated regularly with key database queries (such as the number of transactions currently in transitional states). If counts were building we could detect if certain subsystems were down.

At the same time, we had a page that was updated by software monitoring multiple application logs, looking for certain key phrases.  Many times, if a multiple outage or problem was occurring, we would be notified by at least one of the monitoring methods, many time by both.

The generalized lesson (even beyond software) is that it's good to monitor multiple key indicators that are independently generated.  It increases your chances of catching problems early.

In fact, we found that some messages, if they occurred in a log file were like Spiderman's "Spidey sense!" They indicated that a problem was about to incur in the next 5-10 minutes and allowed us to take preventive measures.

© 2024 Praveen Puri

Notre Dame Cathedral: An Amazing Example of Project Management Success and Strategic Simplicity®

After Notre Dame Cathedral burned down, the French government vowed to rebuild it in 5 years.  Many were doubtful, given the damage to the centuries old structure.

As most of us know, in modern times, major projects almost invariably take longer than planned.  Some giant public-works projects fall behind by decades.

Well, this weekend, after 5 years, Notre Dame opened on time.  This is one of the few massive projects that came in on time.  

One of the reasons for the success was that they followed Strategic Simplicity®.  Like with most projects in the planning stage, people came out of the wood work, suggesting "improvements" and "enhancements" that were now possible with the reconstruction.

The managers, architects, and stakeholders turned them all down.  They focused on one end result: rebuild the beloved cathedral just as it had been.  No scope creep allowed.

© 2024 Praveen Puri

"Billboard Hot 100" for Chatbots

This weekend's WSJ had an article about a pair of grad students at Berkeley who, as part of a project, developed a system to rate chatbots.

Their site, Chatbot Arena, has exploded into one of the most-watched rankings for AI systems.  It works by letting visitors to the site ask a question, which is then answered by two anonymous chatbots.  The visitor then votes for which one answered best. It then tracks the chatbots on a leaderboard.

© 2024 Praveen Puri

A Real-World Example of AI Success

Crain's Chicago Business has just written about a real-world example of business success with AI: Northwest Meat.  This small, family-owned business provides meat to restaurants in Chicago.  They had a daily task that they dreaded: taking orders, and compiling the warehouse schedule for the next day, involving over 1,000 different products.  It involved either the owner, or office staff, spending three hours at night.

Now, they use an AI tool called Choco AI that is designed for food distributors. It's cut their time down from three hours to less than 60 minutes.  Northwest Meat told Crain's that they pay $1100/month for the tool, and the owner estimates it saves him $22k/year in labor expenses.  #AI #distribution #supplychain

© 2024 Praveen Puri