DES303 Week 9: Crit 2 and Reframing Ticker Around Task-Aligned Verification

Introduction

Week 9 was the second major critique point for my DES303 project. In Week 8, I shifted Ticker from a tick-based Focus Integrity system into a broader Business-to-Business (B2B) Artificial Intelligence (AI) senior direction. This pivot came from the Week 7 learning that behavioural tick data alone was not enough to understand what the user was actually trying to do. Ticker could collect app activity, input behaviour, screen changes, camera presence, and system state, but those signals still needed richer task context before they could become meaningful.

For Crit 2, I presented Ticker as an AI senior system. In this version, AI does not only check whether the user is focused. It helps clarify project briefs, create project plans, delegate tasks, generate Standard Operating Procedures (SOPs), and then uses that context to support Focus Integrity. This made the project feel more aligned with the Emerging Technologies brief because it moved beyond individual productivity and began to show a future where AI reshapes collaboration, management, trust, and human freedom. The DES304 Emerging Technologies brief asks students to use emerging technologies and speculative design to reimagine how people in Aotearoa connect, collaborate, and create, while developing speculative artefacts, experiences, devices, systems, or scenarios that provoke reflection on future socio-technical change (Baldwin-Ramult, 2026).

The biggest learning from Crit 2 was that the new direction was clearer, but the gap was still not sharp enough. I was at risk of competing with Motion, competing with to-do apps, and competing with workplace monitoring companies. The critique helped me realise that Ticker should not try to replace those systems. Instead, Ticker should sit across them, import task context from them, and use that context to check task alignment.

Before Crit 2, I was asking:

How can AI plan, delegate, and verify work?

After Crit 2, I started asking:

What happens when AI uses existing work tools to understand the task, then lives beside the worker as a senior who questions, warns, learns, and verifies task alignment?

This week therefore became less about adding features and more about clarifying the project's real gap.

Week 9 was not only a critique week. It was a second ideation point where feedback opened new directions, forced me to compare them, and helped me narrow Ticker into task-aligned verification through a conversational AI senior.

Because Crit 2 was not only a presentation but a test of the project's framing, I needed to document more than the final feedback. I needed to show what I presented, what questions people asked, how I grouped the feedback, what alternative directions I considered, and why I chose the next direction. This is important because DES303 treats the blog as process documentation, not only a record of final outcomes.

The Experience

What I brought to Crit 2

For Crit 2, I presented Ticker as a B2B AI senior system. This was a major shift from the earlier Focus Integrity-only direction. Instead of only showing a score or monitoring dashboard, I showed how Ticker could move from a vague project brief to an operational plan that workers could follow.

The system I presented included:

• a project briefing flow
• an AI clarification question loop
• AI-generated project phases and tasks
• worker roster setup
• AI auto-delegation and assignment rationales
• manager override
• SOP generation
• a connection back to Focus Integrity
• early thinking around Ticker Cube as a possible physical extension

The main user flow was: Briefing → Q&A Loop → AI Plan Streaming → Review & Edit → Auto-Delegation → SOP Generation → Final Approval → Worker Handoff → Focus Integrity.

My intention was to show that Focus Integrity should not judge from rough task titles alone. If the AI knows the task steps, expected tools, expected output, and allowed behaviours, then Focus Integrity can become more task-aware.

FIGURE 1 · CRIT 2 SLIDE DECK EVIDENCE

Open larger view

01Title

Open larger view

02Direction shift

Open larger view

03AI-led work

Open larger view

04Connector layer

Open larger view

05Verification window

Open larger view

06Proof layers

Open larger view

07Ticker Cube

Open larger view

08Ethics

Open larger view

09Crit questions

Feedback I received

The feedback from Crit 2 was useful because it did not only respond to the interface. It challenged the logic of the project. The strongest positive feedback was that the AI senior direction was easier to understand and more aligned with the brief. People could more clearly see the future I was trying to show.

However, the critique also raised several serious issues. The AI senior framing communicates the future more clearly, but if Ticker plans, schedules, delegates, and manages tasks, it risks becoming another AI project-management tool like Motion. The project also needed to explain that Ticker is not only measuring activity, but checking task alignment. Finally, worker consent and human freedom became major ethical sticking points.

Post-Crit Ideation: Possible Ways to Respond

After Crit 2, I did not want to simply defend the version I had presented. The feedback showed that the AI senior direction was clearer, but it also revealed that the project could still move in several different directions. Instead of choosing immediately, I used the feedback as a starting point for another ideation round.

The main critique was that Ticker was at risk of becoming too close to existing categories: AI project-management tools, to-do apps, workplace monitoring platforms, or productivity analytics. That meant the next step was not only to improve the interface. I needed to find a sharper gap.

This follows the Design Council's Double Diamond model, where design moves between divergent thinking and convergent decision-making. The process is not linear: feedback and testing can send the designer back into earlier questions before narrowing again (Design Council, n.d.).

This ideation helped me see that the strongest direction was not to make Ticker better at planning or better at watching. The strongest direction was to focus on the uncomfortable space between those two things:

Ticker should use task context from existing tools, then verify whether human behaviour aligns with that task.

This became the clearer gap: task-aligned verification.

Comparing the Post-Crit Directions

After the brainstorm, I compared the strongest directions against the main critique points. I wanted to choose a direction that was clearer for Capstone, different from existing tools, and ethically stronger.

From this comparison, I decided that the strongest path was not a single product feature. It was a combination of three directions:

1. Task-aligned verification as the core gap.
2. Conversational AI senior as the main interaction model.
3. Workplace simulation as the Capstone experience.

This combination answered the Crit 2 feedback better than the original B2B planning model. It lets Ticker use Motion, Slack, Jira, Canvas, Calendar, or other tools as task context instead of replacing them. It also avoids becoming a raw surveillance platform because the system does not only ask whether the user is active. It asks whether the user's behaviour matches the task context.

Most importantly, the conversational AI senior gives the user a way to respond. Instead of silently judging the worker, Ticker can ask:

Is this still part of your task?

This preserves more human freedom than a strict AI boss model.

How I answered the “Motion” critique

The first major critique was that my B2B planning system could become too similar to Motion. This was fair. Motion already describes itself as an AI-powered work platform that can create projects, generate tasks with deadlines and assignees, build project stages, and reduce manager follow-up work (Motion, n.d.).

If Ticker only generates project plans and assigns tasks, then it is not a strong capstone gap. It is just another AI planning tool. My answer after reflection was that Ticker should not replace Motion. It should use tools like Motion, Slack, Jira, Canvas, and Calendar as context sources.

This gives Ticker a clearer role. Motion asks, “What should I work on, and when?” Ticker asks, “During this focus block, did my observable behaviour match the task I claimed or was assigned to do?” That is a different question.

New post-crit system direction

The clearer system direction is not a replacement for the tools people already use. Ticker should import task context from existing systems, translate that context into a Focus Contract, check behaviour through Focus Integrity, and let the user correct ambiguous judgement before a verified effort summary is created.

How I answered the surveillance critique

The second major critique was that Ticker could become another workplace monitoring system. Normal workplace monitoring often asks: Was the person active? What apps or websites did they use? How much time did they spend working?

Ticker asks: Does the observable behaviour match the task the user claimed or was assigned to do?

That distinction matters. A worker can be active but not aligned with the task. A worker can also look inactive while reading, thinking, planning, or doing physical work. This is why the task context from Week 8 matters. Without task context, Focus Integrity becomes a generic activity score. With task context, it becomes a task-alignment estimate.

However, this does not remove the ethical risk. In some ways, it may make the system more ethically complex. If monitoring becomes more contextual, it may feel fairer and more useful. Because of that, people may accept it more easily. That is exactly the speculative tension I want to show.

How I answered the worker acceptance critique

Another important critique was: Would workers actually want this?

Some people said that they would prefer an AI-mediated system over a human manager watching from behind. An AI system may feel more neutral, especially if it explains why it is asking questions and allows the user to respond.

However, this does not automatically make it ethical. In New Zealand, Employment New Zealand and the Privacy Commissioner outline strict requirements for workplace monitoring, consent, and data storage. This means Ticker cannot be framed as ethically neutral. If used in a workplace, it would need transparency, consent, data minimisation, worker access, correction rights, and clear limits on what managers can see. Ethics is not a side section. Ethics is the project.

How I answered the human freedom critique

If AI chooses the task, generates the SOP, defines the steps, assigns the worker, and checks whether the worker follows the expected process, then the human may stop being a problem-solver.

This made me realise that SOPs should not be treated as fixed commands. They should be treated as context. AI can define the expected work context, but the worker must still choose how to solve the task.

This means the AI senior should not only warn or punish. It should communicate with the user. If the user seems off-task, Ticker should ask: “Is this still part of your task?” rather than immediately marking them as failed.

Ethics risk map opened by Crit 2

The critique made ethics more concrete. The issue was not only that Ticker could watch people. The larger issue was how task verification could become normal if it feels helpful, fair, and polite. I translated the main risks into design responses so the next prototype could show limits, correction, and consent rather than only judgement.

How I answered the Ticker Cube critique

Another critique was about Ticker Cube. If the desktop already has camera and screen monitoring, what is the point of the cube?

I realised that the cube cannot just be another camera. It only makes sense if it supports a different context of work. Ticker Cube should not be my immediate next prototype. I should first make the AI senior layer work on the desktop. After that, Ticker Cube can be used for users who are doing physical labour, studio work, or sketching, where screen data is not enough.

Reflection on Action

Main learning from Crit 2

The biggest learning from Crit 2 was that I needed to make the project more impactful and more precise. The B2B planning system placed Ticker too close to Motion, to-do apps, and workplace monitoring platforms. Ticker should use them as inputs. This preserves the user's existing workflow. The user does not need to abandon Motion or Slack. Ticker reads the task context, negotiates the focus contract, and checks task alignment during the work session.

The post-crit ideation changed how I understood the project. Before Crit 2, I was still trying to make Ticker into a complete AI work system. After the feedback and brainstorm, I realised that completeness is not the goal. Clarity is the goal.

A full AI project-management platform would take the project closer to existing tools like Motion. A full surveillance dashboard would take it closer to workplace monitoring companies. The stronger Capstone direction is more focused: Ticker should show what happens when planning context and behavioural monitoring merge into task-aligned verification.

This also helped me think more like a designer and less like a product builder. My instinct is to build the whole system, but the Crit 2 feedback pushed me to ask what part of the system best communicates the future. The answer is not the full backend or the full B2B workflow. The answer is the moment when the AI senior sees uncertain behaviour, asks the worker to explain, stores the correction, and updates the judgement. That moment contains the whole tension: support, surveillance, freedom, and control.

From AI boss to AI senior

In Week 8, I described the system as AI being the boss. After Crit 2, I realised that “AI boss” may be too harsh if it removes human freedom completely. The stronger version is AI senior.

A senior does not only command. A senior guides, checks, questions, gives feedback, and helps the worker improve. This makes the system more believable and more ethically complex. It is not just a villainous surveillance machine. It is a useful support system that may become controlling because it is useful.

Why this solves the Week 7 failure

In Week 7, the system failed because it watched more than it helped. The new direction changes that: The AI senior does not only observe. It talks.

If the user seems off-task, the system can ask: “You have been on YouTube for 3 minutes. Is this video part of your task, or should I mark this as drift?”

This gives the user more freedom. The user can explain, correct, and shape how Ticker understands their working style. Over time, the system can learn that this user often reads documentation before coding, or uses YouTube for tutorial-based work.

Ethics became the centre

The crit also made me realise that I need to focus more on ethics. The project is interesting because it is ethically uncomfortable. The ethical question is not only: Is surveillance bad? The better question is: What kind of surveillance becomes acceptable when it feels supportive, task-aware, and fair? That is where Ticker becomes a stronger capstone project.

Theory

Task-aligned verification, not AI project management

Crit 2 helped me understand that the project gap needed to become much sharper. In Week 8, I framed Ticker as a Business-to-Business (B2B) AI senior that could clarify briefs, plan work, delegate tasks, generate SOPs, and support Focus Integrity. That direction was clearer than the earlier focus-app direction, but it also created a market problem: if Ticker becomes a full AI project manager, it competes too directly with existing tools.

Motion already positions itself as an AI project-management system that can automate project movement, prioritise team tasks, predict deadlines, balance workload through capacity planning, and give managers project visibility without constant check-ins (Motion, n.d.). ActivTrak already positions itself as a work-intelligence platform that captures behavioural activity, app and website usage, productivity trends, and AI insights across organisations (ActivTrak, n.d.). These precedents made the Crit 2 feedback feel accurate. Ticker should not become another project manager or another activity analytics dashboard.

After Crit 2, the stronger framing is:

Ticker imports task context from existing systems, then uses a conversational AI senior and Focus Integrity layer to check task alignment during a focus block.

This changes the project from “AI project management” to “task-aligned verification”. The difference is small but important. A planner asks what should happen. A monitoring tool asks whether the worker was active. Ticker asks whether the worker's observable behaviour appears aligned with the task context they committed to.

This gave me a stronger post-Crit 2 direction. Ticker should become a layer across existing tools rather than a replacement for them.

Workplace monitoring and privacy

The privacy issue became more central after Crit 2. Ticker cannot be treated as a neutral productivity system because it may collect sensitive work information. It may know what task a person is doing, what app they are using, whether the screen changed, whether they were present, whether they switched context, and whether their behaviour matched the expected task. In a workplace setting, this becomes employee information.

Employment New Zealand states that employers must comply with the Privacy Act 2020 and the Privacy Principles when collecting, storing, using, and sharing employee information (Employment New Zealand, 2025). It also states that workplace monitoring, recording, or filming must be done in line with the Privacy Act 2020 and privacy principles, and that employers should create a workplace policy, consult employees, and make sure employees know why monitoring is needed (Employment New Zealand, 2025). This directly affects Ticker's design. If Ticker is used in Aotearoa, the system needs to make monitoring visible, purposeful, and limited.

The Office of the Privacy Commissioner explains that people have rights to know when information is collected, to have it used and shared appropriately, to have it kept safe, and to access their information (Office of the Privacy Commissioner, n.d.). This changes the design requirements for Ticker. A worker-facing Ticker should not only show a final score. It should show the evidence behind the score, the reason for uncertainty, and the way to correct wrong interpretation.

This means the ethical version of Ticker should include:

This makes ethics central to the project rather than a final warning section. Ticker's design problem is not only “how do I detect focus?” It is “how do I make task verification visible, fair, limited, and contestable?”

Trustworthy AI and explainable uncertainty

The National Institute of Standards and Technology (NIST) AI Risk Management Framework helped me reframe Focus Integrity after Crit 2. NIST describes trustworthy AI through characteristics including validity and reliability, safety, security and resilience, accountability and transparency, explainability and interpretability, privacy enhancement, and fairness (National Institute of Standards and Technology [NIST], 2023). This means Focus Integrity cannot be evaluated only by whether the model gives a score. It also needs to ask whether the score is reliable, understandable, fair, private, and correctable.

This matters because Focus Integrity is not measuring attention directly. It is estimating task alignment through indirect evidence. A user could look inactive while reading or thinking. A user could appear active while avoiding the real task. A browser could be useful or distracting depending on the task. Therefore, the system should not present the Focus Integrity score as truth. It should present it as a confidence-based estimate.

After Crit 2, this gave me a clear design rule:

When Ticker is uncertain, it should ask before it judges.

For example, instead of immediately marking YouTube as distraction, Ticker could ask:

You have been on YouTube for 3 minutes. Is this video part of your current task?

This makes the system more explainable and more human. It also gives the user a chance to correct the system's understanding. Over time, the system can learn that one user often uses YouTube for tutorials, another uses documentation before coding, and another needs low-input reading time before writing. This moves Ticker away from hard surveillance and toward negotiated task alignment.

Human-in-the-loop AI and meaningful oversight

Crit 2 also showed that “human control” cannot be vague. If I say that the worker can still choose how to work, then the interface needs to prove it. LangChain's human-in-the-loop model shows one useful pattern: an agent can pause when a proposed action needs review, and a human can approve, edit, reject, or respond before the action continues (LangChain, n.d.). For Ticker, this suggests that the AI senior should pause at decision points that affect the worker.

Sterz et al. (2024) argue that effective human oversight requires more than a human being present. The human needs causal power, access to relevant information, and the ability to act meaningfully. This matters for Ticker because a simple “I agree” button is not enough. The worker needs enough information to understand the judgement and enough control to correct it.

This changes my next prototype direction. I should not only build a warning system. I need to build a correction system. The warning should open a conversation, not just apply a penalty.

Algorithmic management and AI authority

The AI senior direction also connects to research on algorithmic management. Robert et al. (2020) explain that organisations are increasingly using AI systems to manage workers, but that unfair AI management can reduce effort, increase turnover, and create issues around organisational justice. This is relevant to Ticker because the project sits close to a future where AI is not only a tool, but part of management.

Dong et al. (2025) also show why AI management is ethically complicated. Their experimental study found that workers responded differently to AI management compared with human management, and that AI evaluation could reduce emotional resistance to unfair outcomes. This matters for Ticker because an AI senior may feel more neutral than a human manager. Some workers may prefer being questioned by AI rather than by a person. However, that feeling of neutrality can also make AI authority easier to accept.

This became one of the strongest theoretical insights after Crit 2:

Ticker may be accepted not because it is less controlling, but because its control feels more neutral, useful, and polite.

This makes the project more speculative. It is not only about detecting focus. It is about the future of work culture. What happens when AI becomes a trusted senior because it is always available, always consistent, always polite, and always watching?

API integration as an ethical boundary

After Crit 2, I realised that importing task context from existing tools is not only a technical feature. It is an ethical boundary. If Ticker reads from Motion, Slack, Jira, Canvas, Calendar, GitHub, or other work systems, then it may access sensitive information about tasks, deadlines, communication, team roles, and project status.

OpenAI's tools documentation shows how AI systems can connect to external tools, retrieve files, call functions, or access third-party services through tool configurations (OpenAI, n.d.). For Ticker, this supports the idea of a connector layer. However, the Privacy Act framing means the connector layer should be designed carefully. Ticker should use minimal scopes, clear permission screens, read-only import where possible, and local summarisation before sharing anything with managers.

This creates a more responsible design rule:

Ticker should import task context to support the worker first, not to expand manager surveillance.

This helps me separate Ticker from a workplace-monitoring product. Its default audience should be the worker. Manager-facing reports should be limited, consent-based, and focused on task-level outcomes rather than raw behavioural evidence.

Speculative design and capstone impact

The final theoretical point is speculative design. Dunne and Raby (2013) argue that speculative design can be used to question possible futures rather than simply solve current problems. This is useful for Ticker because the project should not only become a productivity SaaS concept. It needs to make a future condition visible.

The future condition is:

AI becomes accepted as a senior at work because it is useful enough to organise, guide, question, and verify human effort.

This is the main impact of Week 9. Crit 2 showed that the project becomes stronger when it stops trying to be a complete work platform and instead focuses on one uncomfortable future relationship: the worker and the AI senior. The AI senior is helpful because it clarifies tasks and gives feedback. It is also controlling because it defines the task context and checks behaviour against it.

That tension is now the centre of the project. Ticker should not ask only whether people can focus better. It should ask what kind of work culture is created when focus, effort, and task alignment become things that must be verified by AI.

Preparation

Updated Week 10 Plan

After this post-crit brainstorm, my Week 10 plan is clearer. I need to stop expanding the B2B planning tool and focus on the interaction model that makes Ticker feel different.

For Week 10, I will develop:

1. The talking AI senior model. This will test whether Ticker feels more human and less hard-coded when it asks clarification questions instead of immediately warning the user.
2. The Focus Contract layer. This will show what Ticker believes the current task is, what tools are expected, and what behaviours are allowed or risky.
3. The correction and learning loop. This will show how the system learns when the user says “this is related” or “I was actually reading”.
4. The public dashboard concept. This will prepare the Capstone workplace simulation by showing what the audience or manager can see.
5. The Ticker Cube development path. This will explore how the same AI senior could become physical and support room-based work.

The success criteria for Week 10 will be:

• 4/5 viewers understand that Ticker is not another planner.
• 4/5 viewers understand that Ticker checks task alignment.
• 3/5 viewers describe the AI senior as more human than the old warning system.
• 3/5 viewers identify the ethical discomfort without prompting.
• The system asks before punishing uncertain behaviour.
• The Capstone direction feels more like a staged future workplace than a product demo.

Conclusion

Week 9 was important because Crit 2 showed me both the strength and weakness of my new direction. The strength was that the AI senior idea made the project much easier to understand. It showed a clearer future: AI does not only help users work, but begins to organise, guide, question, and verify human work.

The weakness was that I had not yet found the gap clearly enough. Crit 2 helped me realise that Ticker should not compete with Motion or Slack. It should use them. The new direction is to make Ticker a task-alignment layer that imports context and communicates with the user.

This also separates Week 9 from the Week 8 brainstorm. Week 8 opened possible Capstone formats. Week 9 tested those options against Crit 2 feedback and narrowed the project into the strongest direction. Week 10 can now prototype and refine that chosen direction.

Ticker should not become another planner or another surveillance tool. It should become a speculative AI senior that makes task-aligned verification feel useful, human, and ethically uncomfortable.

The next step is to build that conversational AI senior model and test whether it makes Focus Integrity feel smarter, fairer, and more future-facing.

References

• ActivTrak. (n.d.). Work intelligence for productivity optimization. Retrieved May 28, 2026, from https://www.activtrak.com/
• Baldwin-Ramult, L. (2026). DES304: Emerging technologies stream brief [Course brief]. University of Auckland.
• Design Council. (n.d.). Framework for innovation. Retrieved May 28, 2026, from https://www.designcouncil.org.uk/our-resources/framework-for-innovation/
• Dong, M., Brinkmann, L., Sherif, O., Wang, S., Zhang, X., Bonnefon, J.-F., & Rahwan, I. (2025). Experimental evidence that AI-managed workers tolerate lower pay without demotivation. arXiv. https://arxiv.org/abs/2505.21752
• Dunne, A., & Raby, F. (2013). Speculative everything: Design, fiction, and social dreaming. MIT Press.
• Employment New Zealand. (2025, August 14). Employee privacy. New Zealand Government. https://www.employment.govt.nz/fair-work-practices/employee-privacy
• LangChain. (n.d.). Human-in-the-loop. Retrieved May 28, 2026, from https://docs.langchain.com/oss/python/langchain/human-in-the-loop
• Motion. (n.d.). Motion's AI project manager. Retrieved May 28, 2026, from https://www.usemotion.com/features/ai-project-manager
• National Institute of Standards and Technology. (2023). Artificial intelligence risk management framework (AI RMF 1.0) (NIST AI 100-1). U.S. Department of Commerce. https://doi.org/10.6028/NIST.AI.100-1
• Office of the Privacy Commissioner. (n.d.). Privacy Act 2020. Retrieved May 28, 2026, from https://www.privacy.org.nz/privacy-principles/
• OpenAI. (n.d.). Using tools. Retrieved May 28, 2026, from https://platform.openai.com/docs/guides/tools
• Robert, L. P., Pierce, C., Morris, L., Kim, S., & Alahmad, R. (2020). Designing fair AI for managing employees in organizations: A review, critique, and design agenda. arXiv. https://arxiv.org/abs/2002.09054
• Sterz, S., Baum, K., Biewer, S., Hermanns, H., Lauber-Rönsberg, A., Meinel, P., & Langer, M. (2024). On the quest for effectiveness in human oversight: Interdisciplinary perspectives. arXiv. https://arxiv.org/abs/2404.04059