The Two Faces of AI

 

I’m torn…My vocation entails teaching leaders how to solve complex problems both to protect the core business, and to ignite new Growth. My avocation is music - I’m a member of an international band comprising two grizzled veterans & brilliant young players from Europe’s best music schools. In both my vocation & avocation, AI is a blessing and a curse.

Which one will prevail?

AI and My Vocation

AI helps us protect the core business by enabling our management system and helping make problems visible. AI collects data, analyzes value streams, and helps predict hot spots. At the same time, I’m painfully aware that AI agents are shrinking entry level jobs. AI agents are good at core admin tasks & the basic research & analysis new team members used to do. Nowadays, getting that crucial first job is much harder.

Our new Growth work is focused largely on the consumer goods & financial services industries. AI agents enable our sprints. We can run many more experiments & harvest learning quicker.  AI agents also help us analyze & repurpose the existing assets. Years worth of social media content, for example, can be quickly analyzed & reused for new sprints.  We ‘fail fast, fail forward’ with breathless ease.  New assets can be developed & tested while we sleep.

As an engineer, I’m admire & celebrate the splendid engineering & continuous problem solving around neural networks the past twenty years.  Bravo!

I’m also aware that my fellow engineers & I lack the moral & philosophical chops to grapple with the biggest moral challenge of our age: How to deploy the AI genie so as to enhance our lives and reduce suffering while sustaining our core values?

AI and My Avocation

As a musician, I’m acutely aware of AI’s devasting effects on the creator economy. Musicians, writers, journalists, graphic designers, filmmakers and other creators have largely been obsolesced.  Tech companies ‘scrub’ everything they can find and thereby train their ML monsters.  Then they resell us AI slop at twenty bucks a month. Some say we are witnessing the greatest theft of IP in history,

I try to be a good mentor for our young players. One of my core messages is: “AI has destroyed the willingness to pay for creators’ work. We have to accept that hard fact & adapt.  How do we adapt? By becoming ambidextrous.  Develop a second capability that will put food on the table. Be a carpenter, electrician, engineer, teacher, lawyer etc. I do not want you to end up homeless.”

Our band does not use AI to create music – that is unthinkable. We play our instruments with our own hands, we will never use autotune or anything like it. But AI does reduce hassle in the studio. For example, AI can handle the tedious task of eliminating overly loud breathing between vocal phrases.  (I’m a loud breather…).

AI agents help us curate & repurpose our social media assets, as we do in my day job. I engaged a young engineer to develop an AI agent that selects potentially viral content, creates thumbnails & templates for the channels we’re targeting (e.g. YouTube, Instagram, TikTok). AI analyzes a given musical piece’s performance according to our key metrics (Streams, Click-through rate, average watch duration etc.).  Needless to say, I also use him in my day job.

In summary, AI is both a blessing and a curse - in my vocation and avocation.  No wonder I’m torn.

Best wishes,

Pascal Dennis         E: pascal.dennis@leansystems.org

Respect for People in an AI World?

 

Is ‘Respect for People’ possible in the age of AI? A generation ago, the Toyota Production System (TPS) doubled down on this idea and changed the world. Their revolutionary idea was to release the ‘genius of the organization’ through ‘Total Involvement’.  Nowadays, more than a few Tech leaders envision billion-dollar companies with ‘very few people – ideally just one.’ Is ‘Respect for People’ a dead idea?

Middle Management is Over…(?!)

At recent conference a well-known Tech leader opined that, ‘Middle management is over.’ His plan to ‘create value’ amounts to 1) continually replace people with AI agents, and 2) maintain a skeleton crew to ‘manage the work’. 

Avarice & nihilism aside, this hare-brained plan ignores uncomfortable facts, the most obvious being: If we fire everybody, who is going to buy our stuff? Other problems include:

·       Work is messy. You have to deal with internal & external customers, supply chain partners, regulators, banks & insurers, and eccentric, quirky people up & down your value streams and customer journeys. How will you sustain these relationships without people?

·       People are good at building essential relationships – AI agents are not. People mentor younger team members. They know the company’s work better than anyone. As Toyota observed, if you release smart, well-trained people on big problems, they’ll release a tsunami of value.

·       AI agents are expensive and can mess up big-time. The question is not ‘Can an AI agent do this for us?’ The question is: ‘Can an AI agent do it better & cheaper than a human, and with less hassle & risk?

The World is More Than Databases & Loops

‘Software Brain’ seeks to flatten the world and all human experience into a series of databases and loops. Big AI agents can then, presumably, issue instructions and ‘manage the work’.  Here’s the underlying mindset: If the word does not align with the database – change the world. 

In fact, all these things that look like a computer are nothing like a computer. The Law, for example (an early, seemingly easy target for AI agents), is much more than databases and loops. It’s messy, ambiguous and full of grey areas. You can argue different sides and come to different conclusions. The same is true of Medicine, Engineering, Accounting, and Actuarial Science, not to mention the Arts including the visual arts, music, creative writing, theatre and film-making.

So, this thing [Life] that looks like a computer is not at all like computer. Not everything is a loop. Not everything humans experience can be captured in a database. Trying to flatten people & life in this way is the polar opposite of ‘Respect for People’.

People do not consider code-writing to be a blessing. The idea of linking all their databases & automating their lives with Claude Co-work frightens & disgusts them. People hate AI because it flattens them. It's not a marketing problem or a bug - it's a center feature of Software Brain.

TPS remains a sharp rebuke to those who dream of a ‘billion-dollar company of one person’. Is eliminating middle management & destroying entire industries something to brag about?  Big Tech leaders are now routinely booed off the stage. Is it any wonder?

Best wishes,

Pascal Dennis         E: pascal.dennis@leansystems.org

The Trouble with OKRs – part 2

 

Does the OKR approach to strategy deployment reflect the 'top-down' mentality of Big Tech? Does OKR reflect the debilitating phenomenon called ‘Software Brain’, and its core premise - Life is just a series of loops & databases…?

Strategy is a Messy Human Thing

As with many fallacies, ‘Software Brain’ contains a significant element of truth. Life is indeed a ‘series of loops & databases’. Study & understand these loops & databased and you acquire a superpower, which used wisely can greatly reduce human suffering.  But life is much more than this.

But if you try to flatten life in this way you may end up becoming hated, no? Life also comprises messy, unpredictable, thorny, comical things like human beings, their memories, emotions, ideas & intuitions; their dreams, aspirations, religious beliefs and shared conscious & unconscious minds; the cycles & rhythms of nature including human nature, animals & the forces of physics, chemistry, biology, the pure ideas of philosophy and mathematics.

Aristotle described the two Worlds of Necessity and Contingency. Software brain seeks to flatten the world to match algorithms, loops and databases, which is a paltry & incomplete reflection of one of Aristotle’s worlds. Does OKR do the same?

The Essence of Strategy Deployment & Execution

Strategy deployment entails engaging the genius of the organization.  The Leader’s most important job is developing a questioning culture, one in which healthy push-back is encouraged.  “Here’s our Aspiration & Winning Logic.  What do you think? Help us understand what we’re missing; give us fresh perspectives and your ideas on how to get better.”

To be sure, I’ve encountered OKR companies with questioning cultures wherein healthy dissent in encouraged, but they’re a minority.  More common is the superficial, unimaginative Aspiration & Winning Logic, mechanically pushed down with little reflection, testing or validation. People tune out – (‘Here’s one more dumb thing I have to do…’). Do they sense that their humanity is being flattened into loops and databases?

The W-V model attributed to the great Shoji Shiba is a fine example of a healthy Strategy Deployment system. Hypotheses continually bounce back and forth between management levels.  “Here’s what we think? Are we on the right track? If so, please translate our shared thinking into tactics for your team, department or site.”

Most importantly, this back & forth between and among management levels continues into & throughout the strategy execution cycle.

Key Attributes of World Class Strategy Systems

When I’m asked by a Board or Senior Team to ‘assess our strategy & management system’, a key step is to walk the value stream backwards, starting with the customer, and moving upstream one step at a time till you get to raw material suppliers. A parallel step entails walking the customer journey backstream, starting with the top of the proverbial funnel.

In my value stream walks, I ask questions like:

·       Are problems visible (even to an outsider like me)?

·       What do people talk about in their huddles (from senior leaders to the front-line)?

·       What’s the organization’s attitude to problems? Are problems hidden, a sign of failure? Or are they, to quote W. Edwards Deming, ‘the process speaking to us’?

·       Do teams have a standard approach to problem solving?

·       How good is the data? Is it complete, end to end, and of high quality?

·       Does the team develop ‘Your Own Data’ (YODA) as required? Or do they passively wait for corporate data?

These are all messy questions that engage both hemispheres of the brain and both of Aristotle’s worlds. Does your OKR process foster a healthy, pushback culture that can sustain & engage? Or does it flatten corporate life?

Best wishes,

Pascal Dennis         E: pascal.dennis@leansystems.org

 

How Good is Your Management System?

 

What’s the purpose of a ‘management system’? What are its key elements? What are common failure modes?  How do you know if you have a good one?

My job the past few decades has been to help our partners meet their business goals by developing strong people & management systems. My personal mission is to teach leaders Ambidexterity – the ability to both

a) Protect the core with OpEx/Lean, and

2) Ignite new Growth using the methods of Silicon Valley & Singapore.

Your management system enables both and is analogous to the wire harnesses that enable high performance airplanes & automobiles. I’ve written at length about the questions above. In this piece I’d like to describe the tell-tale attribute of a great management system.

Key Attributes of World Class Management Systems

World class management systems make problems visible.

By ‘problem’ I mean the difference between

·       What Should Be Happening, and

·       What is Actually Happening

Each of these must at least define purpose, process & key metrics. By ‘visible’ I mean obvious to all team members.  ‘Everybody knows’, to paraphrase Leonard Cohen.

In practice this means that problems are painfully clear at team boards at every level - from the Senior Leader Obeya to front line team boards & huddle sessions. When I’m asked by a Board or Senior Team to ‘assess our management system’, a key step is to walk the value stream backwards, starting with the customer, and moving upstream one step at a time. A parallel step entails walking the customer journey backstream, starting with the top of the proverbial funnel.

In my value stream walks, I ask questions like:

·       Are problems visible (even to an outsider like me)?

·       What do people talk about in their huddles? Are problems understood & addressed in a clear & decisive way?

·       What’s the organization’s attitude to problems? Are problems hidden, a sign of failure? Or are they, to quote W. Edwards Deming, ‘the process speaking to us’?

·       Do teams have a standard approach to problem solving?

·       How good is the data? Is it complete, end to end, and of high quality?

·       Does the team develop ‘Your Own Data’ (YODA) as required? Or do they passively wait for corporate data?

Here are your key checkpoints:

·       The visibility of problems, and

·       Quality of problem solving (the focus of upcoming articles)

I vividly remember an engagement with a large multinational consumer goods company.  They were good, capable people and their company had enjoyed strong growth for several years. They’d invested in teaching their team members root cause problem solving. I spent a few weeks walking the most important value streams from the customer all the way upstream. I attended team huddles and walked the shop floor.

I presented my findings to the COO & senior management team:

·       Everything looks good on the surface – purpose strategies seem cogent, data seems to be flow, workplace fundamentals look okay on the surface (e.g., 5S, standardized work, visual management), and morale is fairly good.

·       But there's flashing RED light – team boards at all levels are always Green; the critical third wall of the Obeya (Blockers) is usually empty.

‘But we’ve given everybody problem solving training,’ the team responded. ‘

There's flashing RED light,’ I repeated, and added the my old sensei' motto, ‘No problem, is a big problem.’

My report fell on deaf ears.

Six months later there were major outbreaks of food poisoning related to their products. Several customers became ill, and the company’s brand suffered serious damage. A number of senior leaders resigned.

The visibility of problems and quality of problem solving are your early warning system, and a key to de-risking your enterprise.  You want to assess your management system?  Start there.

Best wishes,

Pascal Dennis         E: pascal.dennis@leansystems.org