L’IA pour cadres supérieurs: 6 façons simples de voir l’IA
(article en anglais)
AI for executives: 6 no-nonsense ways to think about AI
(3-4 minute read)
This is the 3rd instalment of Maxa AI’s series of educational content that focuses on the intersection between business performance and machine intelligence.
1. Man or machine?
This famous quote from a leading figure in AI says it all: “It’s not man versus machine; it’s man with machine versus man without. Data and intuition are like horse and rider, and you don’t try to outrun a horse; you ride it.” If you do not start familiarizing yourself with AI as an executive, Darwin will catch up with you (and your business). AI is there to help you focus on true value add. COVID-19 has magnified the distress of firms that neglected low-hanging digital transformation for too long.
2. Prediction machine
AI is not magic. AI is primarily a ‘prediction machine’. This is not always obvious or intuitive. When you forecast customer purchases, when your Tesla steers smoothly on autopilot, or when Siri listens to your requests – these are all machine learning prediction problems. Think of AI where prediction is key.
3. The supermanager with superpowers
Besides its stellar predictive acumen, AI has other related superpowers, such as plowing through and classifying data, and learning new knowledge from it. AI is rigorous: it is allergic to mental shortcuts such as marketing personas. “If AI were a person, with those skills, she would be hired in a New York minute.” Think of AI as the key missing member on your senior team.
4. Framing good questions
Your role as an AI-savvy executive can be simple: ask good questions. Asking good questions implies that you have a good understanding of your business and its processes, you are literate about your data ‘portfolio’ – across finance, sales, customer care, online touchpoints, operations, etc. – and you have intuitions about your blind spots, about where you can do better, or about pockets of data that should be full of signals and insights. Turn over-optimistic AI incantations (“AI is the new thing and will help our business”) into strategic challenges that address clear problems (“Can’t we do a better job of knowing how many red widgets to produce this month?” or “We have a small team: which high-potential customers should they focus their efforts on?”).
5. AI traps
Try to avoid AI traps. Here are a few classics:
- Proper AI training: We hear the horror stories of AI ‘management’ seminars that derail into hyperparameter tuning and the likes. Most AI training is not well suited for execs. Ask around.
- Infrastructure costs: No one wants to be the last guy to buy Blackberry Enterprise Servers. In 2020, it is cheap and straight-forward to store and process data, from multiple messy sources. The day that this becomes an expensive proposition, you know exactly what you are doing and are already generating outsized returns.
- Data scientists: In AI’s early days, ‘getting into AI’ often meant hiring 2-3 data scientists. This does not cut it anymore. You will need a much broader package of expertise, and you may not have the clout to attract seasoned data science, software and other talent in a small non-core practice.
- Scale: As a quick rule of thumb, business performance AI yields best results if you have generated at least $5 million in annual sales, for 3+ years. This is a rough proxy for business scale, data scale, and business process consistency. Online businesses may get away with a lot less.
6. AI as a business initiative, not a tech initiative
Run AI projects like any important business initiative, such as launching a new product line or conducting a major cost-cutting effort. IT managers can be great leaders for AI implementations, but make sure that you have active P&L or function leader sponsorship, and concrete business challenges with short windows for returns. Iterate from there.
About Maxa AI
Maxa AI is the missing machine intelligence that brings together business & sales leadership and traditional FP&A (financial planning and analysis). Our AI platform runs alongside your ERP and transactional data systems to give them superpowers to monitor, forecast and optimize all your business, all the time.
Contact us for a demo on Linkedin or at maxa.ai