Find a worthwhile change.
Map the work and its bottlenecks. Choose a useful starting point and agree how to judge the result—time, quality, and effort.
AI, AUTOMATION + BUSINESS SYSTEMS
Get your team ready for AI, one everyday task at a time. Cut repetitive work and give customers more of your time.
OLD DOG, NEW TECH HOUSTON, TEXAS
X post
“Programming is becoming unrecognizable.”
Forecast
AGI around 2030, with 2029 a possibility.
Interview
“Be an expert in making agents for business.”
Bluesky post
“Novel discoveries will start to come faster than we can absorb”
Mailing list
“AI is a tool, just like other tools we use. And it’s clearly a useful one.”
Forecast
“Soon, they may do the entire task of a software engineer end to end.”
Research
An AI agent’s time horizon measures task difficulty.
Research
AI exceeds human baselines on several benchmarks.
THE LESSON FROM COMPUTERS
A computer in the corner of a paper-run office changes little. The gains come when the team changes how the work moves.
A request goes on a slip. Someone retypes it, checks the file cabinet, and calls a colleague to find out who’s following up.
Log it once. The team can see the customer’s history, who owns the request, and when a reply is due.
AI drafts a reply from approved information and flags missing details. A team member checks it, handles exceptions, and calls the customer when a conversation would help.
Example request: “Can we move our Friday appointment to Monday?”
Draft: “Thanks for letting us know. What time on Monday works for you? I’ll check our calendar and confirm an opening.”
Human check: Confirm availability before promising a time. If the customer has a tight deadline, call to work out the options together.
Illustrative time comparison: If 12 routine requests take five minutes each by hand and two minutes each with AI and review, that’s 36 minutes back for customer calls.
Save paper for the handwritten thank-you. Use the time you get back for the customer who needs you.
A 10-MINUTE EXERCISE
Grab a sheet of paper and pick one task your team does every week. You can do this before choosing an AI tool.
Think of appointment changes, follow-up emails, or copying the same details between systems. Write down what starts the task and who handles it.
Use a made-up example. Mark each step that’s repetitive and each step where someone needs to use judgment.
Time one run today. If you try AI later, compare the total time including review, then check for missed details and errors.
WHAT WE DO
Give your people the skills and systems to use new technology in their everyday work.
Practical sessions built around your team’s real tasks: how to use AI, check its output, handle sensitive information, and know when to use another approach.
Draft a customer reply, check it against company information, and decide what needs a person’s approval.
Start with one workflow. Keep the whole business in view.
HOW WE WORK
Map the work and its bottlenecks. Choose a useful starting point and agree how to judge the result—time, quality, and effort.
Work with your team on a real task. Test the tool, practice using it, and set the access rules and review steps the work requires.
Connect the workflow, document it, and give it a clear owner. Check what improves, fix what doesn’t, and choose where to go next.
WHAT THE RESEARCH SHOWS
Technology adoption is a change in how people work. These studies help explain why we combine tools, training, and thoughtful implementation.
THE LESSON FROM COMPUTERS
“organizational ‘investments’ have a large influence on the value of IT investments”
A review of firm-level evidence connects the value of computing with changes in processes, skills, and organization.
Brynjolfsson & Hitt · Journal of Economic Perspectives, 2000AI IN CUSTOMER SUPPORT
15%
In a study of 5,172 customer-support agents, AI assistance increased this measure by 15% on average. Benefits varied across workers.
Brynjolfsson, Li & Raymond · Quarterly Journal of Economics, 2025CAPABILITY HAS BOUNDARIES
25.1%
In an experiment with 758 consultants, GPT-4 users completed suitable consulting tasks 25.1% more quickly. On a task outside its capability, accuracy fell.
Dell’Acqua et al. · Organization Science, 2026Published findings from specific settings. These are research results, not ODNT client results or forecasts for your business.
Amazon’s 1997 shareholder letter described using the internet to offer a larger selection, easier discovery, and shopping around the clock. It also called for sustained investment in systems and infrastructure.
Our takeaway: ask what the technology lets your business do differently. This is a company’s account of its strategy, rather than a controlled study.
Read the original 1997 letterOLD DOG, NEW TECH
Your team knows its customers, its work, and the problems worth solving. That knowledge is the starting point for learning a new way forward.
Work directly with Adam Peña, founder of Old Dog, New Tech. Teaching, technical direction, and hands-on implementation belong in the same conversation.
EMAIL ADAM
Tell me one task you’d like to make easier. I’ll help you think through what to try, what to measure, and where your team should stay in charge.
adam@odntsolutions.com 713 857 6904