VIEW THIS AS

Auto mode follows the Route Engine until you choose a viewpoint.

YOU ARE HERE

ROUTE CHECK

CONNECTED TO

WHAT NEXT

Use the canonical route for this room, or HELP if you are unsure.

How to Plan Your Week with Super Intelligence

Three secondary students review open books and notes together in a classroom.
Three students studying together with open books at a classroom table.

How to plan your week with Super Intelligence begins with an honest decision about what the week can hold. A useful plan protects real commitments, gives a few meaningful outcomes enough room, and leaves a visible route for change. It does not require every evening to perform like an empty Saturday morning. Its central question is: what can I reasonably agree to do during these seven days, given the life I am actually living?

In this eduKateSG series, Super Intelligence, or SI, is an editorial name for practical AI assistance used with human judgment and repeatable workflows. It is not a claim that current products are proven artificial superintelligence. Here, SI helps sort supplied information, calculate proposed budgets, compare choices and check for conflicts. You decide what matters, verify the result and control any real-world action. None of the examples connects a calendar, sends a message, books an appointment or changes someone else’s plans.

This guide is about the personal week: fixed commitments, household maintenance, chosen projects, relationships, spare capacity and unfinished work. The examples are fictional adult planning packets. Their durations are exercise assumptions, not recommended sleep, work or health prescriptions. You can complete every calculation with paper and a calculator. An assistant is useful only if the combined planning and checking process is worth its time.

The result you are working towards is a short weekly agreement with yourself: a verified capacity budget, a small set of outcomes, a place for recurring maintenance, a reserve that has not already been spent, and a rule for handling disruption. The detailed timing of tomorrow can come later. First, make sure the week itself is feasible.

Choose your reading route

Build your own plan: purpose → inputs → capacity → choices → maintenance.

See the full calculations: Maya’s packet → her plan and alternatives → seven-hour disruption → Daniel’s roster-dependent week.

Keep the system usable: unfinished work → the SI conversation → weekly review.

Practise independently: exercise and complete answer → transfer tests → targeted repairs → questions → your next week.

1. Give the week a clear job

A day plan answers what to do next and when to begin. A weekly plan decides which pieces of life deserve space across several different days. That longer view makes trade-offs visible. You may have enough time for a learning session on Tuesday, but also need a second session before Sunday’s review. You may want to clear a cupboard on Saturday, while knowing that the only quiet window for a personal application is Saturday morning. Both tasks can look reasonable in isolation. Their competition appears when the whole week is considered.

Choose the start and finish of the planning horizon explicitly. Monday to Sunday is convenient for many people, but someone whose work roster arrives on Thursday might plan Friday to Thursday. Label the dates and the local time zone when appointments matter. A seven-day horizon ordinarily contains 168 clock hours, but a local week containing a daylight-saving clock change needs special attention. The worked packets use seven ordinary 24-hour days in one time zone, so their arithmetic is deliberately straightforward.

Give the week a sentence that expresses its main responsibility. For example: “Keep existing commitments, finish the document pack due Friday, and make one contained step on the home project.” That sentence does not erase other parts of life. It explains which optional choices should yield when the week becomes tighter. It is more useful than “be productive”, because it can resolve a real conflict.

Separate the week’s responsibility from your identity. A week in which you maintain essentials while caring for an adult relative may contain little visible project progress. A week with an unusually open weekend may allow a larger creative task. Neither pattern establishes your general ability or worth. Planning works better when the unit under inspection is the arrangement of commitments and resources rather than a judgment about the person living through it.

Start by naming the main uncertainty. Perhaps work hours could change, a delivery window is not confirmed, or a needed document has not arrived. Do not ask SI to pretend that uncertainty is settled so it can produce a tidy calendar. Ask for a plan that shows what is known, what is assumed and which decision will be revisited. A visible conditional choice is often more useful than an unsupported promise.

A practical opening prompt is: “Help me make a seven-day personal plan from the information I supply. First check the horizon, fixed commitments and missing constraints. Do not invent availability or treat tentative plans as confirmed. Show a capacity budget before recommending optional work. Offer alternatives where choices depend on my preferences. Keep all real-world actions as proposals.” Read its response for evidence of understanding before adding more information.

Back to reading routes · Continue: 2. Capture enough reality without exposing your whole life

2. Capture enough reality without exposing your whole life

Capture means gathering the inputs needed to plan, not uploading every message, diary entry or account you own. Begin with your own calendar view, a task list, recent notes and any genuine obligations you have accepted. You can translate these into a small planning packet yourself. “Family commitment, Saturday 14:00–18:00, including travel” is often enough. The assistant usually does not need the person’s name, address, private conversation or reason for the visit.

Record fixed commitments as occupied intervals or reliable totals, and describe what the totals include. A work block of 09:00–18:00 may include a lunch break. If it does, do not later subtract that same break from the remaining hours. A social visit can include travel or exclude travel, but the packet must say which. A weekly plan becomes falsely spacious when travel disappears; it becomes falsely overloaded when the same time is counted twice.

Recurring maintenance belongs in the capture even when it feels too ordinary to deserve a task. Laundry, food preparation, routine cleaning, personal administration, replenishing supplies and maintaining the planning system all consume time. Group genuinely similar small activities when that makes the packet readable. Keep a separate line for a task with a real deadline or unusual consequence. “Household maintenance” should not conceal an important renewal that requires documents before a specific date.

Capture unfinished work with a current state. “Personal essay” is a subject, not an actionable record. A more useful entry says: “Outline complete; first two sections drafted; approximately 90 minutes of editing remains; no external deadline; stored in my usual documents folder.” You can omit the file name from an AI prompt if it is unnecessary. The important information is what exists, what remains, what could block it and whether doing it this week still matters.

For each possible activity, distinguish a commitment from a wish. You may sincerely want to learn a language, organise photographs and repair a shelf. Those desires become weekly commitments only when you choose bounded work that fits. Keep the rest on a possibility list. This prevents a brainstorm from silently becoming an obligation list and lets you appreciate a useful idea without immediately promising to execute it.

Check the packet against reality once. Look for omitted travel, preparation, meals already included elsewhere, repeated tasks under different names and assumptions about someone else’s availability. If the source says “possibly Wednesday”, retain that uncertainty. If two sources disagree, flag the conflict instead of selecting whichever version creates more free time. Capturing less information accurately is often better than collecting a large pile that cannot be checked.

Back to reading routes · Continue: 3. Build a capacity budget with non-overlapping categories

3. Build a capacity budget with non-overlapping categories

The first calculation is a boundary, not a productivity target. Start with the clock hours in the chosen horizon. Subtract occupied time using categories that do not overlap. Then divide the remainder between recurring maintenance, protected personal time, chosen project work and unallocated reserve. Every hour should have at most one accounting label. A category can contain several activities, but it cannot also be hidden inside another total.

Do not force your life into another person’s budget. Use your own known needs and commitments, including recovery, accessibility requirements, care responsibilities and realistic transitions. The numbers in this article show a method rather than a desirable lifestyle. If your occupied time is larger than a worked example’s, the useful conclusion is a smaller discretionary commitment, not pressure to cut an essential need until your figures resemble the example.

There are two tests after the total balances. The first is placement: do the remaining hours occur before the relevant deadlines? Four free hours on Sunday do not complete an obligation due on Friday. The second is suitability: are those hours usable for the activity? A shared journey might allow quiet reading but not confidential administration. Two fragments of twenty minutes may help with sorting receipts but not with an uninterrupted practical task requiring setup.

Label candidate windows with simple constraints instead of pretending to measure your mind precisely. “Quiet desk access”, “out of home”, “interruptible”, “only after another commitment” and “uncertain finish time” can be enough. Your experience may show that some windows are unreliable for concentrated work. Preserve that observation without turning it into a permanent diagnosis. A weekly plan should be able to accommodate variation in attention and circumstances.

Keep reserve visible. Reserve is time left uncommitted so it can absorb a change. It is different from leisure you have deliberately protected and different from an optional task you are already expecting to do. If a planner calls two hours “buffer” and simultaneously assigns a two-hour project there, the buffer has been spent. An optional list can offer choices for unused reserve later, but those choices are not part of the base commitment.

An arithmetic check is necessary but insufficient. Ask SI to show the additions, then verify them independently with a calculator. Ask it to name any overlap in category definitions and any deadline that depends on time occurring too late. OpenAI’s guidance on whether ChatGPT tells the truth warns that responses can be inaccurate, including confident responses. A polished table therefore deserves the same calculation and source checks as an informal answer.

Back to reading routes · Continue: 4. Choose outcomes and say what will wait

4. Choose outcomes and say what will wait

Weekly priority is a choice between feasible alternatives. Begin with real obligations and their consequences. Then identify one or a few optional outcomes that would make the week worthwhile to you. The right number depends on their size, the available windows and the amount of maintenance already required. There is no universal rule that every adult needs three major goals each week. One well-chosen outcome can be plenty.

Write an outcome as a visible end state. “Work on photographs” is vague. “Select twenty photographs and place them in one labelled album, without redesigning the full archive” gives a boundary. “Improve my home” could consume an unlimited number of hours. “Measure the cupboard, compare two storage arrangements and choose whether to proceed” is a contained decision. The boundary helps you estimate, stop and recognise completion.

A useful priority explanation includes the cost of postponement and the cost of doing it now. Suppose the photograph album is a gift for a visit next weekend, while the cupboard has no deadline. Completing the album may deserve this week’s quiet window. But if the album requires photographs that another person has not supplied, promising completion could be unreasonable. You might instead complete the selection you control and record the missing input as a condition.

Ask SI for two defensible plans rather than a single allegedly optimal answer. One plan might protect a learning milestone and postpone a household improvement. Another might finish the household improvement and preserve learning at a small maintenance level. Have it explain what each choice gains, delays and risks. Then choose using your values and circumstances. A numerical ranking can support the conversation, but it cannot decide how much a relationship, rest or personal interest should matter to you.

Make the “not this week” list explicit and short. It is a record of decisions, not a catalogue of failures. State the reason and the next reconsideration condition. “Archive older photographs after the current album is complete” is clearer than a date you do not believe. “Reconsider the course after the roster is confirmed” preserves a dependency. Some items can be removed entirely when they no longer matter. Keeping a stale aspiration forever does not protect it.

Finally, check whether the chosen priorities fit the resources they actually require. A three-hour project might need a borrowed tool, a quiet room or another adult’s participation. Those resources are not automatically available because your calendar is blank. Do the controllable preparation first, keep unconfirmed coordination conditional, and decide what independent work can replace a blocked task. This is how a weekly plan becomes more robust without becoming much longer.

Back to reading routes · Continue: 5. Keep maintenance and projects from crowding each other out

5. Keep maintenance and projects from crowding each other out

Maintenance restores the conditions that allow life to continue; project work creates a chosen new result. Both matter, but they behave differently in a weekly plan. Maintenance often returns after completion. Finishing this week’s laundry does not eliminate future laundry. A bounded project, such as completing a document pack, should eventually leave the active list. Treating every recurring task as a permanent project can make the plan look endlessly unfinished.

Give recurring work a workable service level for this particular week. “Clean the whole home perfectly” is not the same commitment as “complete the routine kitchen and bathroom clean”. If you shorten a task, say what is excluded and whether the reduced version is still acceptable. Never reduce a necessary safety check merely because a schedule is crowded. For consequential maintenance, follow the applicable instructions or obtain appropriate help rather than letting an assistant improvise a shortcut.

Bundle only when the activities genuinely fit together. Combining an errand with an existing journey may reduce additional travel, but only if the location, opening time and carrying requirements work. Do not assume a shop is open or a service is available without checking. In the fictional packets, those conditions are supplied as assumptions. In your own plan, distinguish a confirmed fact from a possibility worth investigating.

Put an upper bound on active personal projects. You can keep many interests without actively advancing all of them during the same week. A small active set makes it easier to protect enough time for completion and to resume after an interruption. The useful test is whether every active project has a credible next step and a suitable window. If several have neither, they are possibilities being described as commitments.

Budget the overhead of the planning method itself. Capturing information, asking an assistant, reading its output, verifying calculations and transferring decisions into your chosen system all take time. That time should appear somewhere in the maintenance budget. A plan that saves fifteen minutes but requires an hour of weekly formatting is not automatically an improvement. Keep the readable version short enough that you will actually refer to it.

For a recurring task, use previous observations as a starting estimate when they are relevant. If recent laundry sessions took longer because the drying arrangement changed, include that changed condition. Do not average away an important difference just to make the number smaller. For a new task, use a bounded first step or a range and leave room for uncertainty. A clear estimate with a stated assumption is more useful than an exact-looking number without a basis.

Back to reading routes · Continue: 6. Worked packet: Maya's ordinary week

6. Worked packet: Maya’s ordinary week

Maya is a fictional adult who works a regular weekday schedule and lives independently. Her planning horizon is Monday through Sunday in one time zone, with no clock change. She wants to finish a personal document pack due Friday at 23:00, maintain a learning routine and make progress on a photograph album. All information below is supplied for the exercise. The plan is not based on a real person’s calendar, and none of the amounts is a recommendation for another person’s needs.

Her occupied-time packet has five non-overlapping categories. She reserves 56 hours for sleep across the week, using eight hours on each example day. Her weekday work blocks total 45 hours, nine hours per day including the lunch interval. Work travel totals five hours, one per weekday. Meals outside those work blocks, personal care and ordinary transitions total 21 hours, three per day. Two fixed weekend social commitments occupy six hours in total, three on Saturday and three on Sunday, with their travel already included.

The first check is 56 + 45 + 5 + 21 + 6 = 133 occupied hours. Subtracting 133 from 168 leaves 35 hours. That number is not all project capacity. Maya still needs household maintenance, protected personal time and reserve. Her available-window summary is three hours on each weekday and ten hours on each weekend day: 3 + 3 + 3 + 3 + 3 + 10 + 10 = 35. The daily and weekly totals agree.

For weekday placement, the supplied routine is sleep from 23:00 to 07:00, work from 09:00 to 18:00, and travel from 08:30 to 09:00 and 18:00 to 18:30. The three routine hours outside work are 07:00–08:30, 18:30–19:30 and 22:30–23:00. This leaves 19:30–22:30 as the three-hour candidate window. Its existence does not mean she intends to concentrate for three hours every evening. It is simply the interval available for the remaining categories.

On Saturday and Sunday, the exercise assumes the ten available hours on each day are distributed around the supplied routine and social commitment. Maya has checked that each weekend day includes a suitable uninterrupted four-hour project window. That placement condition matters: a total of ten scattered hours alone would not prove the project arrangement possible. When using your own packet, write down the actual windows rather than borrowing this assumption.

Maya allocates seven hours to recurring maintenance, eight to protected personal time, twelve to chosen projects and eight to reserve. The second check is 7 + 8 + 12 + 8 = 35. The full-week check is 133 + 35 = 168. Protected time is Maya’s chosen leisure and recovery allowance within the remainder. It is not a claim about the amount anyone medically needs, and it is not spare capacity waiting for a project to consume it.

Her maintenance definition includes one hour of laundry, two hours of routine cleaning, two hours of food preparation and shopping beyond the ordinary meal time already counted, one hour of personal administration, and one hour of planning and review. These add to seven. The separation prevents meals, shopping and food preparation from being counted under two different labels. Her actual household arrangement might use different categories; what matters in the exercise is that the definitions are explicit and mutually exclusive.

Back to reading routes · Continue: 7. Worked result: a feasible allocation and two genuine alternatives

7. Worked result: a feasible allocation and two genuine alternatives

Maya’s document pack requires four hours: one hour to gather the supplied documents and check the list, ninety minutes to prepare the missing explanatory notes, and ninety minutes for final review and organisation. The exercise assumes no additional document must arrive from someone else. “Done” means the complete pack is ready for Maya’s own final submission decision before the Friday 23:00 deadline. The plan does not submit anything. Monday’s one-hour project allocation and the Tuesday and Thursday ninety-minute allocations supply the four hours before the deadline.

The learning outcome is three hours of practice across two weekend sessions. Maya will complete a supplied lesson exercise and independently explain the method afterwards. The photograph-album outcome is five hours: two and a half hours on Saturday to select and order the photographs, followed by two and a half hours on Sunday to assemble and check the album. She is not committing to reorganise the entire photo collection. The three chosen outcomes therefore require 4 + 3 + 5 = 12 hours.

Here is the complete allocation of the 35-hour remainder, with each day’s categories in the order maintenance, protected time, projects and reserve. Monday is 1, 1, 1 and 0 hours. Tuesday is 0, 1, 1.5 and 0.5. Wednesday is 1, 1, 0 and 1. Thursday is 0, 1, 1.5 and 0.5. Friday is 1, 1, 0 and 1. Saturday is 2, 1.5, 4 and 2.5. Sunday is 2, 1.5, 4 and 2.5. Each weekday sums to three; each weekend day sums to ten.

Reading down the same allocation gives the weekly checks. Maintenance is 1 + 0 + 1 + 0 + 1 + 2 + 2 = 7. Protected time is 1 + 1 + 1 + 1 + 1 + 1.5 + 1.5 = 8. Projects are 1 + 1.5 + 0 + 1.5 + 0 + 4 + 4 = 12. Reserve is 0 + 0.5 + 1 + 0.5 + 1 + 2.5 + 2.5 = 8. This is a feasible allocation under the supplied assumptions, not evidence that the activities have happened.

The weekend project blocks each combine ninety minutes of learning and two and a half hours of album work. Maya can place those within the verified four-hour window or separate them if suitable windows exist. The weekly plan does not need to dictate every minute. It needs to preserve the total, the conditions and the relationship between the two album stages. If Saturday’s selection is incomplete, Sunday cannot simply proceed as though that dependency has been satisfied.

Alternative A protects a quieter weekend. Maya keeps the four-hour document pack and three hours of learning, but chooses a two-hour photograph-selection step instead of completing the five-hour album. Projects fall from twelve to nine hours. She explicitly adds the released three hours to protected time, taking it from eight to eleven, while reserve remains eight. The remaining budget is 7 + 11 + 9 + 8 = 35. The cost is that the album will not be complete this week; the gain is a less demanding weekend.

Alternative B gives more attention to the album. Maya keeps the four-hour document pack and five-hour album, reduces learning to a one-hour continuity session, and assigns the released two hours to reserve. The budget becomes seven maintenance, eight protected, ten project and ten reserve hours. It still totals 35. This choice is reasonable only if a smaller learning step is acceptable to Maya; SI should not quietly shrink an external course requirement or claim that one hour is equivalent to three.

Maya chooses the original plan because the supplied week is relatively predictable and the two weekend windows have been checked. Her record includes both rejected alternatives and their reasons. If circumstances change, she does not have to invent new options under pressure. A good assistant can preserve those alternatives in a few sentences without turning the weekly plan into a long dossier.

Back to reading routes · Continue: 8. Worked repair: seven hours disappear during the week

8. Worked repair: seven hours disappear during the week

Now change the packet. On Tuesday morning, Maya learns that work will occupy two additional hours on Tuesday and two on Thursday. A separate necessary household issue will consume three hours of Saturday’s previously available time. These are fictional changes, accepted as fixed for this exercise. Their combined effect is seven fewer hours in the discretionary remainder. The occupied total rises from 133 to 140, leaving 28 rather than 35. The original arrangement no longer fits its original windows.

The first trap is to say, “There were eight hours of reserve, so nothing needs to change.” The arithmetic works globally, but the placement does not. Tuesday and Thursday now each have only one available hour. Each already contains an hour of protected personal time, so their planned ninety-minute document sessions cannot remain there. Saturday falls from ten available hours to seven. A reserve located on Sunday cannot help a document pack that needs to be ready by Friday night.

Maya preserves the four-hour document outcome and the eight hours of protected time. She also retains the seven-hour maintenance allowance. To make the repair comfortable enough to operate, she reduces the album outcome from five hours to two hours of selection only and moves the three-hour learning outcome to next week’s consideration list. Project work is now six hours: four for the document pack and two for photograph selection. The new reserve is seven hours because 28 − 7 − 8 − 6 = 7.

This is a six-hour reduction in project commitments alongside a seven-hour loss of available time. The original reserve was eight hours and the repaired reserve is seven. The accounting therefore reconciles: seven lost hours are absorbed by six fewer project hours and one fewer reserve hour. Nothing has been funded by quietly reducing sleep, compressing the fixed social visits, double-booking protected time or assuming that AI makes a four-hour task take an hour.

The repaired allocation, again maintenance, protected time, projects and reserve, is Monday 1, 1, 1, 0; Tuesday 0, 1, 0, 0; Wednesday 0, 1, 2, 0; Thursday 0, 1, 0, 0; and Friday 1, 1, 1, 0. Saturday is 2, 1.5, 2 and 1.5, totalling seven. Sunday is 3, 1.5, 0 and 5.5, totalling ten. The weekday totals are now three, one, three, one and three. Together with seven and ten at the weekend, they equal 28.

The document pack receives Monday’s one hour, Wednesday’s two hours and Friday’s one hour. Friday’s final check can occupy 19:30–20:30, safely before the supplied 23:00 deadline, with Maya retaining control of any submission. Wednesday uses its former one-hour reserve plus the hour of maintenance moved to Sunday. Friday uses its former one-hour reserve. Tuesday and Thursday each retain their protected hour. The sequence of gathering, preparation and checking is preserved even though the session lengths have changed.

Read down the repaired allocation. Maintenance is 1 + 0 + 0 + 0 + 1 + 2 + 3 = 7. Protected time is still eight. Project work is 1 + 0 + 2 + 0 + 1 + 2 + 0 = 6. Reserve is 1.5 + 5.5 = 7. The total is 7 + 8 + 6 + 7 = 28, and 140 occupied hours plus 28 remaining hours equals 168. This checks the arithmetic, placement and scope of the repair together.

There is a real cost: the document pack loses its earlier completion margin and is now ready on the deadline day. Maya should acknowledge that reduced margin rather than describing the revised plan as equally robust. If Friday’s window becomes doubtful, the next decision is to reduce another flexible commitment or renegotiate the deadline where possible. A smaller album and postponed learning are explicit choices, not invisible failures. The plan has adapted to changed conditions while preserving a truthful account of what remains feasible.

Back to reading routes · Continue: 9. A second packet: Daniel's roster-dependent week

9. A second packet: Daniel’s roster-dependent week

Daniel is another fictional adult. His work days vary, and a roster change may be announced midweek. He shares a home with another adult, but this packet includes only obligations he has personally accepted. The example avoids assuming that someone else can absorb a task just because Daniel is busy. His horizon is again an ordinary seven-day, 168-hour week. It includes 56 hours reserved for sleep, 40 hours in confirmed work blocks, six hours of work travel, 21 hours of personal routine and nine hours of fixed personal commitments including their travel.

Those categories total 56 + 40 + 6 + 21 + 9 = 132, leaving 36 hours. His verified remaining windows are four hours Monday, two Tuesday, eight Wednesday, two Thursday, four Friday, eight Saturday and eight Sunday. The sum is 36. Wednesday’s larger window comes from the supplied roster; it is not an assumption that everyone has a midweek day free. Only Wednesday and Saturday contain uninterrupted three-hour windows suitable for his main personal project.

Daniel’s maintenance budget is nine hours. It includes the household tasks he owns, shopping, administration and planning. Protected personal time is ten hours. His chosen project outcome is an eight-hour document-and-photo inventory for a planned move, with the limited aim of recording what is in three specified storage areas and identifying missing information. It excludes packing, arranging transport, making payments and negotiating with a landlord. He retains nine hours of reserve. The budget is 9 + 10 + 8 + 9 = 36.

His baseline allocation, in the same four-category order, is Monday 1, 1.5, 1, 0.5; Tuesday 0.5, 1, 0, 0.5; Wednesday 2, 2, 3, 1; Thursday 0.5, 1, 0, 0.5; Friday 1, 1.5, 1, 0.5; Saturday 2, 1.5, 3, 1.5; and Sunday 2, 1.5, 0, 4.5. Each day fits its supplied window. Summed by category, the allocation gives nine maintenance hours, ten protected hours, eight project hours and nine reserve hours.

The project sequence matters. Monday’s hour establishes the inventory categories and gathers existing notes. Wednesday’s three hours cover the first two storage areas. Friday’s hour checks gaps and prepares the final area. Saturday’s three hours complete that area and review the inventory. Sunday has no promised project work. Its reserve can absorb an overrun, but the base outcome is intended to be complete without relying on that reserve.

The uncertain event is a possible four-hour additional work commitment on Wednesday, which would reduce that day’s remaining window from eight hours to four. Daniel does not schedule the extra work as both confirmed and unconfirmed. He keeps a baseline and a conditional branch. The trigger is the actual roster notice. Until it arrives, the Wednesday project block is provisional; no dependent promise is made to another person on the assumption that it will happen.

If the change occurs, the remainder falls from 36 to 32 hours. Wednesday becomes one hour of maintenance, one of protected time, one of project work and one of reserve. This releases one maintenance hour, one protected hour and two project hours from that day. Daniel moves the maintenance and protected hours to Sunday, reducing Sunday’s reserve from 4.5 to 2.5 hours. The two project hours are removed by shrinking this week’s outcome to the first two storage areas, with the third explicitly deferred.

Under this smaller scope, Monday still establishes the categories, Wednesday begins the first area, Friday checks gaps, and Saturday finishes the two-area inventory and reviews it. The repaired category totals are therefore nine maintenance hours, ten protected hours, six project hours and seven reserve hours. They add to 32. The logic is different from Maya’s repair: Daniel reduces the scope of one project rather than suspending a separate learning outcome. He still completes a coherent inventory of two areas. SI should preserve that changed definition of completion rather than later reporting the original three-area project as finished.

A valid alternative is to keep the eight-hour project and use two of Sunday’s remaining reserve hours for the third area, but only after verifying that Sunday’s space, access and concentration conditions support the work. Under that alternative, reserve falls to five hours and the total remains 32. Daniel chooses the smaller scope because the roster uncertainty has already materialised once and he prefers to retain more room. The reason is personal and explicit; an assistant cannot derive it from arithmetic alone.

Back to reading routes · Continue: 10. Carry unfinished work forward without selling the same hour twice

10. Carry unfinished work forward without selling the same hour twice

At the end of a week, do not copy every unchecked item into the next plan automatically. An unfinished item needs a decision. It may still be important, may need a different next step, may be blocked, or may no longer deserve space. Copying it without checking the new week’s capacity turns yesterday’s estimate into tomorrow’s obligation. The new week already has its own fixed commitments and recurring maintenance.

Use one continuing record for each meaningful outcome. Give it a short recognisable label, the current state, the remaining work, its reason for being active and the next review condition. Maya’s revised photograph project might read: “Album: selection session completed; chosen images are in one folder; ordering and assembly remain; estimate three hours; reconsider at next weekly planning.” That wording preserves progress. It does not recreate the original five-hour task or pretend that selection alone completed the album.

The next week’s budget should contain only the remaining effort that is actually chosen. If Maya has three hours of album work left and accepts it into the new week, she reserves three suitable hours once. She does not also retain a five-hour “make album” entry under personal projects and a separate three-hour “finish photographs” entry in an overflow list. Those are two descriptions of the same work. SI can help identify likely duplicates, but Maya confirms whether the records truly refer to the same outcome.

Separate waiting from working. Suppose the album now needs two photographs from another adult. The wait itself does not consume three hours of active work. It does create uncertainty about whether an assembly session will be usable. Record the dependency and a check point. If the photographs do not arrive by the point when the session must be committed, choose an independent task or leave the window uncommitted. Do not count both possible tasks as guaranteed completions.

Be careful with deferred recurring work. A missed weekly clean might mean next week’s job takes longer, or it might mean a reduced clean is still appropriate. The answer depends on the task and conditions, not a universal rule. Record any accumulated work honestly and inspect the new estimate. Conversely, do not automatically double the next maintenance allowance merely because one occurrence was skipped. The physical state of the task determines the work, not the number of unchecked boxes.

A carry-forward note should also preserve decisions that prevent scope from growing again. If Daniel chose to inventory two storage areas, the third area remains a separate future choice. It should not reappear in an assistant’s summary as an overdue obligation unless Daniel has actually made that commitment. A useful request is: “Reconcile these records. Preserve completed work and the current scope. Flag duplicate descriptions and uncertain estimates. Do not add deferred possibilities to next week’s commitments until I choose them.”

Before closing the week, ask one practical question about each active outcome: can I resume it without rereading the whole conversation? If the answer is no, add the next physical or mental action and the location of the necessary material in your own system. A short resumption note can save more effort than a long motivational summary. It also lets you switch tools, take a break or plan without AI while keeping continuity.

Back to reading routes · Continue: 11. Use SI through a controlled weekly conversation

11. Use SI through a controlled weekly conversation

The Life series follows Capture → Clarify → Expand → Structure → Execute → Review → Improve. For weekly planning, each stage has a distinct output. Capture produces the small factual packet. Clarify resolves ambiguous deadlines, overlapping categories and uncertain commitments. Expand explores a few possible arrangements. Structure produces the checked budget and chosen outcomes. Execute means the person carries out the plan, or separately authorises an appropriate action. Review compares the plan with what occurred. Improve changes the next planning attempt using that evidence.

The sequence prevents an attractive answer from skipping a necessary question. If the assistant starts by allocating project hours before it knows whether travel is included, return to Clarify. If it proposes only one plan despite a genuine trade-off, return to Expand. If its total is wrong, return to Structure. If an outcome is marked complete because a session was scheduled, return to Review and ask what evidence actually shows completion.

Ask, Think, Build, Integrate and Operate describe another useful set of roles. Ask establishes the question and missing information. Think compares the consequences of the available choices. Build creates a usable weekly brief. Integrate places that brief into the calendar or task system you already use, under your control. Operate keeps the record current as the week changes. These roles do not require five applications or increasing levels of automation. You can perform them with one document and a short conversation.

Begin with a constraint-checking pass: “Here is my packet. List contradictions, double counts and missing placement information. Distinguish confirmed inputs from assumptions. Do not create a plan yet.” Respond only to the questions that affect feasibility or important choices. If the assistant asks for unnecessary private details, provide a less identifying description. A generic occupied block is usually enough to prevent a collision.

Then request the budget: “Calculate the remaining hours. Show every category total and the daily-window sum. Keep protected time, maintenance, project work and reserve separate. Identify deadlines that cannot use later windows.” You should be able to reproduce its calculations. If you cannot understand where a number came from, the budget is not ready to guide commitments. Ask for the simplest arithmetic rather than a more elaborate visual presentation.

For the choice pass, request two plans with reasons: “Offer one plan that protects this week’s main outcome and one that reduces load. For each, state what is completed, what waits and what risk remains. Do not imply either option is objectively best.” After choosing, ask for a concise final brief. It should contain the dates, fixed constraints, selected outcomes, allocation totals, conditional branches, deferred items and next review point. Keep the full calculations available, but avoid carrying every explanatory paragraph into the operating view.

At the boundary to action, distinguish a proposal from a change. An assistant describing a calendar entry does not prove that an entry exists. A tool-enabled assistant may have permissions that differ from a chat-only tool, so verify the specific capability and authorisation before relying on it. The method in this guide works without connected tools. If you choose to integrate it, review the exact target, dates, attendees and consequences before any external invitation, booking or other consequential action.

Back to reading routes · Continue: 12. Review the week using observations rather than guilt

12. Review the week using observations rather than guilt

A weekly review asks what the plan should learn. Start with facts: which outcomes reached their stated completion condition, which remain partly done, which were deliberately removed, and which changed because their assumptions changed. Separate those states. A task cancelled after a sensible reconsideration is not the same as a task forgotten despite a still-valid commitment. A completed two-area inventory is not an incomplete three-area inventory when the scope was explicitly changed before execution.

Compare estimates with actual effort only where the comparison is meaningful. If Maya expected five hours for an album but deliberately did only selection, comparing two actual hours with five estimated hours does not show a three-hour efficiency gain. Compare the selection step with its own estimate, or record that the scope changed. Likewise, extra time caused by a missing input should not automatically be attributed to poor concentration. Keep the cause connected to the evidence.

Review reserve by location as well as total. Maya’s first plan had eight hours of reserve, but much of it was at the weekend. Her disruption showed that deadline-sensitive work needed earlier options. The lesson is not necessarily “add more buffer everywhere”. It may be “finish this kind of document earlier”, “hold a Wednesday alternative”, or “keep the Friday deadline clear in the brief”. A targeted repair addresses the mechanism that actually caused difficulty.

Look for invisible costs of using SI. How long did it take to prepare the packet, correct assumptions and transfer the plan? Did the assistant introduce tasks you did not need? Did its repeated suggestions make it harder to keep a chosen boundary? A useful system can be very simple. If one manual calculation and a short list are clearer than an AI-generated dashboard, use the simpler arrangement.

Choose a small improvement for the next week. You might redefine one maintenance category, add travel to a recurring appointment, shorten the active project list or use a range for an unfamiliar task. Avoid redesigning the entire system after one unusual week. First ask whether the event was a one-off disruption or evidence of a recurring mismatch. More observations may be needed before changing a default estimate.

For a modest personal trial, compare several planning cycles using the same basic questions: did important commitments remain visible, were selected outcomes feasible, was the plan easy to adjust, and did planning take a reasonable amount of effort? You are looking for a useful pattern in your circumstances, not proving that AI causes a universal productivity improvement. No guaranteed gain follows from using an assistant, and a truthful negative result is valuable if it helps you simplify.

Back to reading routes · Continue: 13. Independent practice: repair an overloaded packet

13. Independent practice: repair an overloaded packet

Try this exercise before reading its answer. A fictional adult, Aisha, has a seven-day horizon with 138 hours already occupied in non-overlapping categories. The remaining windows are two hours Monday, two Tuesday, four Wednesday, two Thursday, four Friday, eight Saturday and eight Sunday. She needs six hours of recurring maintenance and chooses eight hours of protected personal time. Her proposed projects are a four-hour application pack due Friday at 20:00, a five-hour home archive with no deadline and a six-hour learning assignment she has independently chosen. She wants at least five hours of reserve.

First calculate whether the request fits. Then propose two complete alternatives without reducing occupied time, maintenance, protected time or the five-hour reserve floor. For one alternative, prioritise the application and learning assignment. For the other, prioritise the application and archive. Explain which optional work is deferred or reduced. Finally, show a feasible daily allocation for one alternative and identify the extra information required to be sure the deadline can be met.

The total remainder is 168 − 138 = 30 hours. The daily windows agree because 2 + 2 + 4 + 2 + 4 + 8 + 8 = 30. The proposed remainder allocation is 6 + 8 + 4 + 5 + 6 + 5 = 34 hours. It exceeds capacity by four hours. The correct diagnosis is an overcommitment under the stated constraints. Reformatting the calendar cannot fix it, and describing all three projects as high priority does not change the arithmetic.

Alternative one keeps the application at four hours and the learning assignment at six. It reduces the archive to a one-hour preliminary sorting step, assuming that step is useful and separable. Project work totals eleven hours. With six maintenance, eight protected and five reserve hours, the total is 30. The original five-hour archive is not complete; four hours remain outside this week’s commitment. If the archive cannot be usefully split, defer it entirely and increase reserve or protected time by the unused hour.

Alternative two keeps the four-hour application and five-hour archive, while reducing the self-chosen learning task to a two-hour introductory step. Projects again total eleven hours, so the same category totals fit. This is valid because the exercise explicitly makes the learning assignment self-chosen. If it were an externally required six-hour assignment with an unchangeable deadline, the reduction would not be an equivalent solution. That changed fact would require reconsidering the archive or another flexible commitment.

For alternative one, use the category order maintenance, protected time, projects and reserve. Monday is 0.5, 1, 0.5, 0. Tuesday is 0.5, 1, 0.5, 0. Wednesday is 0.5, 1, 2, 0.5. Thursday is 0, 1, 1, 0. Friday is 1, 1, 0, 2. Saturday is 1.5, 1.5, 4, 1. Sunday is 2, 1.5, 3, 1.5. These rows total two, two, four, two, four, eight and eight hours respectively.

The column totals are six maintenance, eight protected, eleven project and five reserve hours. Application work uses the half-hour slots on Monday and Tuesday, two hours Wednesday and one hour Thursday, totalling four before Friday. The weekend project allocation is six hours of learning and one of archive sorting. For example, Saturday can hold three learning hours and the archive hour, while Sunday holds the other three learning hours. Neither weekend task is being used to fund the earlier application deadline.

The remaining placement question is whether those application slots occur at suitable times with the necessary materials and whether its steps can be split as proposed. If preparation requires an uninterrupted two-hour window, Wednesday supplies one in the allocation only if the actual window is contiguous. If an input arrives Thursday afternoon, the earlier preparation assumptions may fail. A correct total is therefore an answer to one part of the exercise; a reliable plan also verifies sequence, access and timing.

Back to reading routes · Continue: 14. Transfer tests: use the method when the numbers change

14. Transfer tests: use the method when the numbers change

Transfer means applying the reasoning to a different constraint instead of copying the example’s hours. In the first test, take Aisha’s chosen alternative one. On Thursday morning she loses two hours from Friday’s remaining window because of a newly confirmed commitment. The application is still due Friday at 20:00. What changes, and what should stay unchanged? Do not move work unnecessarily just because the plan has a new version.

Friday’s two-hour reserve can absorb the lost hours arithmetically, but doing only that violates the five-hour reserve floor carried forward from the original exercise. Its remaining window falls from four hours to two, occupied by one maintenance hour and one protected hour. The application remains scheduled through Thursday, so the new Friday commitment does not displace its allocated work. Weekly reserve falls from five hours to three, while the other category totals remain six, eight and eleven. These total 28, matching the new remainder. To preserve the stated floor, Aisha must reduce optional project work by two hours. Accepting only three reserve hours would require a new, explicit decision to relax the original constraint; it is not a valid solution under the unchanged packet.

A valid repair under the original constraints is to defer the one-hour archive step and reduce the self-chosen learning step by one hour, restoring five reserve hours. The repaired budget is six maintenance hours, eight protected hours, nine project hours and five reserve hours, totalling 28. The learning outcome must be renamed to match its five-hour scope. The assistant should not declare the original assignment complete or quietly weaken the constraint.

In the second test, a task takes ninety minutes but can only happen before a service closes at 17:00. The week contains six unallocated evening hours and two unallocated hours on Sunday. Can the assistant declare that there is enough capacity? The answer is no. The totals show enough time in general but provide no compatible window. You need an actual pre-17:00 opportunity, a verified alternative service arrangement or a changed commitment. More evening reserve does not solve a daytime access constraint.

In the third test, an assistant proposes an eight-hour project, then lists a three-hour first stage as a separate priority. The eight-hour estimate already includes that stage. What should happen? Keep one eight-hour total, with the three-hour stage shown as a component rather than an additional charge. If the stage is the only work chosen this week, allocate three hours and defer the remaining five. A clear parent-and-component relationship prevents double-booking without losing useful detail.

In the fourth test, a project is described as “half done” after three hours of effort, but the hardest stage remains and has never been attempted. Can you estimate the remainder as another three hours? Not from that statement alone. Ask what concrete parts are complete, what remains and which assumptions support the next estimate. You can plan a bounded investigation step rather than commit to full completion. Percentage labels are less useful than a visible state when stages differ in difficulty.

In the fifth test, your week is already full before optional projects are added. The useful answer is not to insist on a weekly growth goal. Check for double counts, distinguish fixed from renegotiable commitments and make any necessary trade-off openly. If essentials genuinely occupy the available capacity, a maintenance-focused week may be the truthful plan. SI can help prepare options and questions, but it cannot create time by choosing a more optimistic tone.

Back to reading routes · Continue: 15. Diagnose common failures and make the smallest useful repair

15. Diagnose common failures and make the smallest useful repair

If every week begins with the same oversized list, inspect the selection step. You may be treating captured possibilities as accepted commitments. Keep the capture list, but require each active outcome to have a reason, a boundary and a suitable allocation. Move the rest to a clearly non-committed place. The repair is a stronger choice at the start of the week, not a more insistent reminder at the end.

If the budget fits but tasks repeatedly collide, inspect category definitions and actual windows. Travel may be missing, work may overlap a personal commitment, or two people may need the same room at the same time. Correct the source information before adjusting estimates. A planning assistant can detect inconsistencies in the packet it receives, but it cannot guarantee that the packet includes all the constraints of your life.

If a project repeatedly expands while you work on it, inspect the completion condition. “Organise everything” may hide several independent outcomes. Choose the smallest useful finished state and name what is outside it. For the album, selection, assembly and a complete archive redesign are different scopes. When a new improvement occurs to you, capture it as a possible next project rather than silently adding it to the current week’s promise.

If reserve is always consumed by predictable work, change the baseline. Regularly occurring administration belongs in maintenance, even if its exact content changes each week. A recurring work overrun may require a larger occupied-time estimate or a discussion about the commitment. Reserve remains useful for variation, but describing known demand as a surprise prevents the plan from learning. Adjust the category that explains the repeated observation.

If the assistant keeps rescheduling the same item without asking whether it matters, strengthen the carry-forward decision. Require a reason for continued activity and a current remaining-effort estimate. You may decide to finish it, reduce it, defer it to a meaningful condition or remove it. The quality of that decision matters more than maintaining an uninterrupted chain of planned dates. Repeatedly moving an unwanted task creates administration without preserving value.

If the system becomes exhausting, reduce its resolution. You may only need the fixed commitments, three or four category totals, one chosen outcome and the next review point. Do not maintain minute-by-minute records merely because a tool can produce them. Keep detail where a deadline, dependency or safety issue needs it. A personal weekly plan should make your choices easier to understand and execute, rather than becoming another demanding project.

Back to reading routes · Continue: 16. Questions people ask about planning a week with SI

16. Questions people ask about planning a week with SI

Should I plan every hour of the week?

No. Account for the main demands so the total is credible, then use the level of detail the situation needs. Fixed appointments require timing. A deadline-sensitive project may require a protected window. Leisure can remain open, and reserve should remain genuinely uncommitted. The worked packets use complete arithmetic to teach the method; your everyday brief can be much shorter once you understand the categories and have checked the important constraints.

How much reserve should I keep?

Choose an initial amount based on the variability you actually face, then review whether it was useful. This guide does not establish a universal percentage. A week with a stable roster and few dependencies differs from one with uncertain work hours and several unconfirmed arrangements. Also consider where reserve sits. A large Sunday reserve may do little for a Friday deadline. Keep protected personal time separate so an apparent buffer is not simply rest relabelled as available work.

Can SI tell me which personal priorities are correct?

It can help make consequences, assumptions and alternatives easier to see. You remain responsible for the value judgment. An assistant does not know the importance of a relationship or personal commitment merely from a task label. State what you want protected and ask it to explain the cost of its proposal. If two plans are both feasible, choosing the one you prefer is a legitimate decision without an invented productivity score to justify it.

What if I do not know how long a task will take?

Use a small investigation step, a range or a clearly stated provisional estimate. Define what you will learn before committing the rest of the week. For example, spend a bounded session inspecting the documents needed for a personal application, then estimate preparation after seeing the requirements. Include checking and organisation in the scope where appropriate. Do not let a generated estimate become a promise to someone else simply because it is written confidently.

Should unfinished work always take priority next week?

No. Age is information, but it does not automatically determine value. An unfinished obligation with a genuine approaching deadline may deserve priority. An optional project you no longer want may deserve removal. Review why the work remains, what is already complete and what it would displace now. Carry forward the current state and only the remaining work you choose. This keeps old plans from controlling the new week by default.

What if my week changes every day?

Use a shorter reliable commitment window inside the weekly horizon. Keep confirmed obligations, minimum maintenance and a small set of eligible next steps. Treat later work as conditional until the relevant information arrives. Review when a meaningful constraint changes rather than rewriting the entire plan after every minor interruption. Daniel’s roster packet shows how a baseline and one explicit branch can preserve clarity without pretending that the whole week is predictable.

Can I use the method without sharing my calendar with an AI service?

Yes. Create an anonymous packet of occupied totals, candidate windows and task descriptions. Keep names, addresses, private messages and unnecessary sensitive details out of it. You can also ask for a blank reasoning checklist and perform the arithmetic yourself. If you later decide to connect a service, inspect its actual permissions and privacy choices. The practical method does not require broad access or automatic actions.

How do I know whether the method is helping?

Look for fewer hidden conflicts, more realistic commitments, clearer decisions about unfinished work and less effort needed to recover from a changed week. Include the time spent preparing and checking the AI output. A plan that produces more text but no better choices may need simplifying. Your evidence is the usefulness of the complete process in your circumstances, not the attractiveness of one generated schedule.

Back to reading routes · Continue: 17. Make your next weekly agreement

17. Make your next weekly agreement

For your first attempt, choose one ordinary upcoming week and gather only the information needed to plan it. Confirm the horizon and fixed commitments. Define overlapping categories carefully enough to avoid counting the same hour twice. Calculate what remains, then place maintenance and protected personal time before adding optional projects. Choose a bounded outcome that fits a real window, and retain reserve whose location is useful for the uncertainties you face.

Next, write the decision that gives the plan meaning. Name the outcome you will protect, the optional work you will leave for later and the condition that would trigger a revision. Make sure unfinished work has a current state rather than a copied label. Keep an alternative ready for the most plausible disruption. You do not need to predict everything; you need a way to notice when the original assumptions no longer apply.

At the end of the week, inspect what happened without rewriting the past. Preserve genuine completions, acknowledge changed scope and record the remaining work. Choose one small adjustment supported by what you observed. A useful weekly system develops through these clear, ordinary decisions. Super Intelligence can assist with the comparison and checking, while the week remains yours to live.

Continue with the Life Super Intelligence hub for the broader personal system. Use How to Manage Your Time with Super Intelligence to examine commitments and capacity more broadly. For an organisational project involving teams, handoffs and shared accountability, the separate planning and task management guide covers that workplace context.

The practical standard is simple: your weekly plan should tell the truth about capacity, make a few choices clear, and remain understandable when life changes.

Back to reading routes