Workplace Bot Habits

BotStalling

Theme: Resistance

BotStalling is the habit of holding back. People avoid AI, or keep putting off real adoption, even when it could clearly help. Work carries on the way it always has, and the tools sit largely unused.

BotStalling sits at the origin, low on both effectiveness and efficiency. It is the starting point, not a destination.

What this behaviour looks like in the workplace?

• “Let us wait and see” becomes the default answer to any new tool.
• Manual processes continue unchanged, even where AI could remove obvious drudgery.
• Pilots and trials launch with enthusiasm, then quietly never reach production.

Why this happens?

Resistance is rarely about capability. It is usually fear, uncertainty, or low trust: worry about getting it wrong, about looking foolish, or about what AI might mean for the role. When fear outweighs the appetite to experiment, people stay put.

What’s the cost of this behaviour?

The organisation captures none of the value it is paying for. The tools sit idle, the manual workload stays exactly where it was, and teams that keep stalling fall further behind those quietly building real habits.

How can organisations progress from this state?

Progress starts with confidence and permission, not pressure. A few principles help a stalling team move off the starting line:

• Make it safe to experiment. Give explicit permission to try AI on low-risk, real work, and treat early stumbles as learning rather than failure.
• Start with the work, not the tool. Pick one genuine friction point people already feel, and let AI earn trust by solving it.
• Let peers lead. Adoption spreads faster when a trusted colleague shows the way than when it is pushed from the top.
• Make leadership use visible. When leaders use AI openly, hesitation drops and permission starts to feel real.

BotSitting

Theme: Micromanagement

BotSitting is the habit of over-supervising AI. People do use the tools, but they hover: checking, re-checking, and re-prompting until the time AI was meant to save is spent watching it instead.

BotSitting sits to the lower right of the map, effective but not efficient. Quality holds up, yet the promised time savings quietly disappear into supervision.

What this behaviour looks like in the workplace?

• Rewriting the same prompt again and again in search of a perfect answer.
• Rechecking every line of output, even on low-stakes work.
• Finishing a task barely faster than doing it the manual way.

Why this happens?

Over-supervision usually comes from low trust in the output paired with high personal accountability. It is also diligence with no off-switch: when there is no agreed standard for what “good enough” looks like, people check everything just to be safe.

What’s the cost of this behaviour?

The time AI saves is reabsorbed into watching it, so real capacity is barely freed. People tire of babysitting the tool, and the time savings the business paid for never actually materialise.

The wrong counter-move: BotSwitching.

The tempting fix is to blame the tool and switch to another, then another, hoping one will finally need less supervision. That is BotSwitching. Hopping between tools adds switching cost without building the trust or standards that actually reduce checking, so it keeps a team busy inside BotSitting rather than moving it towards BotSyncing.

How can organisations progress from this state?

The shift is from checking everything to checking what matters:

• Set clear quality bars. Agree what “good enough” looks like so review effort matches the stakes.
• Match verification to risk. Reserve deep checking for high-stakes work and lighten it elsewhere.
• Build trust through evidence. Track where AI is reliable and where it is not, so confidence is earned rather than assumed.
• Protect the core, delegate the rest. Keep human judgement on the decisions that matter and let AI carry the routine load.

BotShitting

Theme: Mismanagement

BotShitting is the habit of trusting AI too much. People submits what the tool produces without really reading it, understanding it, or being able to defend it. Speed goes up, and so does risk.

BotShitting sits to the upper left of the map, efficient but not effective. Work moves fast, but polished output hides errors that surface downstream.

What this behaviour looks like in the workplace?

• Pasting AI output straight into work without reading them
• A hallucinated figure or fact reaching a customer.
• “The AI said so” standing in for human judgement.
• Confidence in a slick-looking answer that nobody has actually checked.

Why this happens?

When output looks finished, the usual warning signs disappear, so people stop looking closely. Deadline pressure, fatigue from over-checking elsewhere, and the quiet assumption that someone else will catch it all push work out unverified. Counterintuitively, the more capable the tool, the easier it is to fall into this.

What’s the cost of this behaviour?

Small errors compound into poor decisions, and mistakes reach customers under the organisation’s name. Accountability blurs, trust erodes, and the clean-up costs far more than the check would have.

The wrong counter-move: BotSlaying.

The tempting fix is to overreact and shut AI down. After unverified output causes damage, leaders ban the tools or clamp down so hard that people stop using AI altogether. That is BotSlaying: “killing” the bots instead of fixing the trust and verification gap. It throws the value out with the risk and pushes the team back towards resistance, never towards BotSyncing.

How can organisations progress from this state?

The shift is from blind trust to accountable use:

• Keep a human in the loop. Make a named person responsible for what goes out, especially on higher-stakes work.
• Verify what matters. Build light checks into the workflow so speed does not skip the substance.
• Reintroduce friction on purpose. Slow down at the decision points where a confident-sounding error would be expensive.
• Reward honesty over polish. Make it safe to say “I have not checked this yet” rather than pass work off as done.

BotSyncing

Theme: Moderation

BotSyncing is the healthy centre. Humans and AI each do what they are best at, in rhythm: AI drafts, summarises, and spots patterns at speed, while people supply judgement, context, and accountability.

BotSyncing sits on the healthy diagonal, where effectiveness and efficiency rise together. It is the habit every other one is trying, and often failing, to reach.

What this behaviour looks like in the workplace?

• AI produces the first draft or the analysis, and a person decides what actually matters.
• Checking is matched to the stakes, not applied to everything.
• A named human stays accountable for whatever goes out.
• People know when to reach for AI and when to leave it out.

Why this works?

Neither over-checking nor blind trust wins. BotSyncing keeps a human in the loop for the judgement machines lack, while letting AI carry the volume that slows people down. The friction that remains is productive, because it is where quality control and real learning happen.

What’s the payoff?

Time saved is genuinely freed rather than reabsorbed, and quality holds because judgement stays human. Teams in this habit report both higher productivity and higher quality, the combination the research calls high AI achievers.

How can organisations sustain and build on this?

The move from here is to spread it, not to relax it:

• Keep judgement human. Protect the decisions and craft that make the work yours.
• Hold the standard. Keep quality bars and clear accountability in place as volume grows.
• Capture what works. Write down the prompts, checks, and workflows that are paying off, ready to share.

BotScaling

Theme: Cultivation

BotScaling is the habit of spreading what works. Once a team has found a healthy way of working with AI, that pattern is documented and adopted by others, so the gains compound beyond a single team.

BotScaling sits further up the healthy diagonal. Effectiveness and efficiency are now rising across teams, not just individuals.

What this behaviour looks like in the workplace?

• A workflow proven in one team is written up and handed to others.
• Shared prompts, checks, and guardrails replace everyone starting from scratch.
• Cross-functional peers, not only leaders, champion what works.

Why this works?

Value stops being trapped in a few individuals. Adoption spreads fastest peer to peer, so proven patterns travel further and faster than any mandate could push them.

What’s the payoff?

Individual wins become organisational gains, closing the gap between people who feel more productive and a business that actually performs better.

How can organisations sustain and build on this?

• Scale only what is genuinely working. Do not spread a habit that has not reached BotSyncing first.
• Give patterns an owner. Someone should keep shared workflows current as tools change.
• Make success visible. Recognise the teams and peers whose ways of working others choose to copy.

BotSoaring

Theme: Maturity

BotSoaring is the habit becoming the culture. Healthy AI use is no longer something to prompt or police, it is simply how the organisation works.

BotSoaring sits at the top right of the map, high on both effectiveness and efficiency, and sustained over time rather than in bursts.

What this behaviour looks like in the workplace?

• Teams instinctively know when to use AI and when to leave it out.
• Good habits persist without reminders or crackdowns.
• New tools are absorbed without the organisation sliding back into old failure modes.

Why this works?

Excellence holds because it rests on habits and judgement, not on any single tool. When the norm is healthy, new people inherit it and new technology is folded in without drama.

What’s the payoff?

AI compounds into a durable advantage rather than a passing spike. The organisation keeps its gains as tools, teams, and tasks change.

How can organisations sustain and build on this?

• Keep learning live. Treat AI fluency as ongoing, not a one-off rollout.
• Watch for drift. Even mature teams can slip towards over-checking or blind trust under pressure.
• Reinvest the dividend. Put the time AI frees into higher-value work and new capability, not just more output.