AI changes how we produce
By 2034, the amount produced before price changes is 36.4 percent higher than in 2024. Your output-per-hour setting helps drive this result; more joblessness holds growth back.
You’re reading the Plain English version. The detailed version explains the technical terms and equations.
Could AI help everyone live better?
Explore what might happen to jobs, prices, and people's spending power as AI spreads—and how yearly payments might help. These are examples of possible futures, not predictions.
Starting with actual 2024 numbers. Government data gives us the starting point. Future effects of AI and payment plans are guesses. See the starting numbers and checks. Output before price changes uses 2017 dollars for comparison. Prices across the whole economy are measured separately from prices consumers pay.
Your story: Faster change · 10 years · Payments that adjust · Government borrowingYour settings stay the same as you move between tabs. Every result after 2024 is an example, not a forecast.
YOUR CURRENT EXAMPLE · 2024–2034
More produced. A better life for everyone?
Jobs and prices shape daily life
Joblessness ends at 6.0%—2.0 percentage points higher than the starting rate. Consumer prices are 29.4 percent higher over the whole period. These are different changes: the final year's price-rise rate is 2.4%.
Choices affect how gains are shared
The selected plan pays $2,055 per adult in the final year, costing $0.55 trillion that year before taxes and other costs. Compared with no payments under the same assumptions, the rough spending-power score is 2.1 points higher and the final yearly price-rise rate is 0.2 percentage points higher.
What this means
These sentences are calculated from your current settings. They describe this example, not a prediction. The model does not yet show how different families are affected.The rough spending-power score ends at 109.6: 9.6 percent higher than its starting value of 100. This is a group-level example, not a promise about any family's budget. Payments do not directly raise real output here, and the funding choice does not calculate actual tax bills or debt.
Does the spending-power score improve?
This rough score brings jobs, output per hour, payments, and prices together. It starts at 100 and does not describe every family's experience.
Rough spending-power score (2024 starts at 100)Start 100.0 → End 109.6
What happens if your assumptions change?
Choose a starting story, then change a slider. The charts and the story in the first tab follow your choices. Try one change at a time.
Output before price changes (trillions, 2017 dollars)Output before price changes (trillions, 2017 dollars)The amount the economy produces, with price changes taken out. Think of this as comparing the amount of goods and services, rather than bigger price tags. Values are in trillions of dollars using 2017 prices as a reference. One trillion is 1,000 billion. The model starts with official 2024 data, then uses your assumptions. Giving people payments does not directly increase this output in this model. These are possibilities, not forecasts.
31.80Output at each year's prices (trillions)Output at each year's prices (trillions)The economy's total output valued at each year's prices, in trillions of dollars. This can rise because more is produced, because prices rise, or both. It is worked out from output before price rises and a separate economy-wide price measure. That price measure is different from the consumer-price measure used for everyday shopping.
48.84Yearly consumer price rise (%)Yearly consumer price rise (%)How much consumer prices rise or fall in one year, as a percentage. At 3%, something costing $100 would cost about $103 a year later. A fall from 3% to 2% means prices still rise, just more slowly. A negative number means prices fall. The starting rate uses official annual averages; later rates are based on your assumptions. The model limits them to −8% through 15%.
2.40People out of work (%)People out of work (%)The share of people in the workforce who do not have a job and are looking for one. At 5%, that is about 5 in every 100 workers. The starting numbers come from official data. Your automation and job settings change the later results. The workforce size stays fixed, and the rate is kept between 1% and 30%. This is a scenario, not a forecast of actual unemployment.
6.04Yearly payment per adult ($)Yearly payment per adult ($)The payment each adult gets in the last year shown. It is dollars per person per year—not per month, per household, or the total over all years. The model assumes every US resident aged 18 or older gets it. Payments are limited to $30,000 a year. Rising prices reduce what the money buys. Taxes and changes to other benefits are not included.
2,055Yearly cost of payments (trillions)Yearly cost of payments (trillions)The total cost of that year's adult payments, in trillions of dollars. It multiplies the payment per adult by the fixed number of adults. Paying $10,000 each would cost about $2.67 trillion a year. This leaves out administration, taxes collected, debt interest, and savings from replacing other benefits. It is not the government's overall budget shortfall or the cost of all years added together.
0.55Rough spending-power score (starts at 100)Rough spending-power score (starts at 100)A rough spending-power score that starts at 100. A score of 110 means this model's score has risen by 10%. It combines jobs, output per hour, payments, and consumer prices. It uses an assumed $55,000 yearly earnings amount to compare payments with work income. It is not a real household budget and does not mean every family is 10% better off.
109.61How things change over time How things change over timeThese charts show how the same scenario changes year by year. Each chart has its own scale, so a steeper-looking line is not necessarily a bigger change. The first point is actual 2024 starting data; later points are imagined results. Start and End give the numbers. The model has not worked out how likely these paths are.
Each chart shows the same story. Each has its own scale, so compare its numbers rather than how steep the line looks.
Output before price changes (trillions, 2017 dollars)Start 23.3 → End 31.8
Yearly consumer price rise (%)Start 2.9 → End 2.4
People out of work (%)Start 4.0 → End 6.0
Yearly payment per adult ($)Start 0.0 → End 2,055.4
Output-per-hour score (2024 starts at 100)Start 100.0 → End 141.1
Rough spending-power score (2024 starts at 100)Start 100.0 → End 109.6
See the year-by-year numbers
Year by year Year by yearA year-by-year list. The first row is 2024 starting data; later rows are simulated, not later actual data. Output is in trillions of dollars. One output column removes price changes; the other uses each year's prices. Price rises and joblessness are percentages. Payments are yearly dollars per adult. Spending power is a rough score starting at 100.
| Year | Output before price changes (trillions, 2017 dollars) | Output at each year's prices (trillions) | Yearly consumer price rise | People out of work | Yearly payment per adult ($) | Rough spending-power score |
|---|---|---|---|---|---|---|
| 2024 | 23.3 | 29.2 | 2.9% | 4.0% | 0 | 100.0 |
| 2025 | 24.1 | 30.9 | 2.9% | 4.2% | 0 | 100.4 |
| 2026 | 25.0 | 32.7 | 2.8% | 4.4% | 0 | 100.9 |
| 2027 | 25.8 | 34.5 | 2.7% | 4.6% | 0 | 101.4 |
| 2028 | 26.6 | 36.4 | 2.7% | 4.8% | 153 | 102.3 |
| 2029 | 27.5 | 38.4 | 2.6% | 5.0% | 434 | 103.4 |
| 2030 | 28.3 | 40.3 | 2.6% | 5.2% | 749 | 104.6 |
| 2031 | 29.2 | 42.4 | 2.5% | 5.4% | 1,073 | 105.8 |
| 2032 | 30.1 | 44.5 | 2.5% | 5.6% | 1,400 | 107.0 |
| 2033 | 30.9 | 46.6 | 2.4% | 5.8% | 1,728 | 108.3 |
| 2034 | 31.8 | 48.8 | 2.4% | 6.0% | 2,055 | 109.6 |
Which assumptions make the biggest difference?
Which assumptions matter most? Which assumptions matter most?Which sliders make the biggest difference to the final spending-power score? We move one slider a little below and above your setting while keeping everything else fixed. ‘A little’ means 10% of that slider's full range, stopping at its limits. Bigger numbers mean a bigger difference in this check. This is not a measure of how certain the result is. Zero may mean that slider does not apply to your chosen payment plan.
We nudge each slider up and down by 10% of its full range, one at a time. Bigger scores mean a bigger change in the final spending-power score.
More output per hour
24.79
Cost savings slowing price rises
2.26
How widely automation spreads
1.40
Pressure on jobs
1.39
Target for yearly price rises
1.12
How strongly payments react
1.01
New jobs balancing losses
0.92
How quickly payments adjust
0.11
Try a separate example of everyday costs
Could everyday essentials get cheaper? Could everyday essentials get cheaper?A separate example of everyday costs, starting at a score of 100. Lower means the basket gets cheaper: 90 means about 10% cheaper in this example. We assume how much producers' savings reach shoppers and how shortages limit those savings. These settings are guesses. This calculator is separate: changing it does not change the main output, price-rise, or payment results.
Everyday-cost score: 93.59 (a separate example; starts at 100)
housing: cheaper productionHow much cheaper might this category become to produce over the whole period? This is not a yearly price fall or an automatic discount at the shop. Only some savings are assumed to reach shoppers, and shortages can hold them back. For example, 20% cheaper production with half passed on and a 40% blockage gives a 6% lower shopper cost. This changes only the separate everyday-cost score. Share of this shopping example: 34.0%; share of savings reaching shoppers: 0.4; share blocked by shortages: 0.6.
energy: cheaper productionHow much cheaper might this category become to produce over the whole period? This is not a yearly price fall or an automatic discount at the shop. Only some savings are assumed to reach shoppers, and shortages can hold them back. For example, 20% cheaper production with half passed on and a 40% blockage gives a 6% lower shopper cost. This changes only the separate everyday-cost score. Share of this shopping example: 8.0%; share of savings reaching shoppers: 0.7; share blocked by shortages: 0.3.
food: cheaper productionHow much cheaper might this category become to produce over the whole period? This is not a yearly price fall or an automatic discount at the shop. Only some savings are assumed to reach shoppers, and shortages can hold them back. For example, 20% cheaper production with half passed on and a 40% blockage gives a 6% lower shopper cost. This changes only the separate everyday-cost score. Share of this shopping example: 14.0%; share of savings reaching shoppers: 0.6; share blocked by shortages: 0.4.
healthcare: cheaper productionHow much cheaper might this category become to produce over the whole period? This is not a yearly price fall or an automatic discount at the shop. Only some savings are assumed to reach shoppers, and shortages can hold them back. For example, 20% cheaper production with half passed on and a 40% blockage gives a 6% lower shopper cost. This changes only the separate everyday-cost score. Share of this shopping example: 25.0%; share of savings reaching shoppers: 0.3; share blocked by shortages: 0.6.
transportation: cheaper productionHow much cheaper might this category become to produce over the whole period? This is not a yearly price fall or an automatic discount at the shop. Only some savings are assumed to reach shoppers, and shortages can hold them back. For example, 20% cheaper production with half passed on and a 40% blockage gives a 6% lower shopper cost. This changes only the separate everyday-cost score. Share of this shopping example: 19.0%; share of savings reaching shoppers: 0.6; share blocked by shortages: 0.3.
How do payment plans change the outcome?
Each row uses the same assumptions and number of years. Choose a plan to update the story, charts, and map. Costs are yearly payments before tax receipts, interest, and other costs.
YOUR CURRENT EXAMPLE · 2024–2034
What does your payment choice change?
The selected plan pays $2,055 per adult in the final year, costing $0.55 trillion that year before taxes and other costs. Compared with no payments under the same assumptions, the rough spending-power score is 2.1 points higher and the final yearly price-rise rate is 0.2 percentage points higher.
What this means
These sentences are calculated from your current settings. They describe this example, not a prediction. The model does not yet show how different families are affected.The rough spending-power score ends at 109.6: 9.6 percent higher than its starting value of 100. This is a group-level example, not a promise about any family's budget. Payments do not directly raise real output here, and the funding choice does not calculate actual tax bills or debt.
Compare payment plans Compare payment plansTry all four payment plans with the same settings and compare their final year. Differences come from the model's rules, not proof that one plan is best. Payments do not directly change real output here, so that number may be the same for every plan even when prices and payments differ. The cost column shows payments only, before taxes and interest.
| Payment plan | Output before price changes (trillions, 2017 dollars) | Yearly consumer price rise | Yearly payment per adult ($) | Yearly payment cost (trillions) | Rough spending-power score | Try this plan |
|---|---|---|---|---|---|---|
| No payments | 31.8 | 2.2% | 0 | 0.00 | 107.5 | |
| Same payment each year | 31.8 | 3.2% | 10,000 | 2.67 | 110.0 | |
| Payments that adjust | 31.8 | 2.4% | 2,055 | 0.55 | 109.6 | |
| Payments linked to output per hour | 31.8 | 2.6% | 3,901 | 1.04 | 110.8 |
Why is output the same across these plans? This model does not give payments a direct effect on real output. The rows do not prove one policy is best, and we do not yet show who pays the taxes or which households gain.
Why did the result change? Choose a box, then read one of its links. You can look at different years without changing your assumptions.
How your choices connect Explore how the choices connectChoose a box to light up its links. Choose a link below to read what it means. The numbers follow your settings and the year you choose. Solid lines show arithmetic; dashed lines show assumed effects. Gray links do not apply to your payment plan. The adjusting plan uses last year's price rise to set this year's payment. This map explains this example, not proven laws about the real world.
Choose a box to see what affects it and what it affects. Then choose a link below to learn why. Numbers follow your settings. These arrows show the rules of this example; they do not prove how the real world works.
Year shown in the mapMove this slider to look at a different year. It does not change your story or settings. The first year, 2024, uses actual starting data and pays no adult payment. Later years are imagined. A link may apply to your plan but still have no effect when a setting or payment is zero.
━━ How totals add up┄┄ A link we assume↶ Uses last year's resultGray: this plan does not use this link
Yearly payment per adult
Your payment plan decides how many dollars each adult gets each year. The adjusting plan changes payments gradually using the speed slider. Payments affect prices, the total payment bill, and the rough spending-power score.
At 2025: $0 / adult / year.
The separate everyday-cost example and missing links
The five everyday-cost sliders change only this separate shopping-cost example. Its current score is 93.59. Cheaper production → some savings reach shoppers, with shortages holding some back → an overall shopping-cost score. This does not change the main price or payment results.
Some real-world links are missing here. Automation does not automatically increase output per hour. Payments do not directly raise real output. The payment bill does not change taxes or debt. The workforce does not grow. We do not show how gains are shared between families. We leave those arrows out because this model does not calculate them.
What do we know, and what are we assuming?
The starting numbers come from government data. Future links between AI, jobs, prices, and payments are assumptions. Matching old numbers does not prove the future story will happen.
How this example works How this example worksThis explains the rules and what is missing. Official data sets the starting point. The links between AI, jobs, prices, and payments are simplified assumptions. Correct arithmetic and working software do not prove the future results will happen. The source section shows which numbers are observed and which choices are guesses.
Where we stand: Official past numbers set the starting point, and one known year's growth is matched. Future AI effects and payment results are still assumptions. See the sources, dates, and limits of the checks.
How the totals add up: The dollar value of the economy combines how much it produces with economy-wide prices. The yearly payment bill is the payment per adult times 267.455 million adults. Joblessness starts close to the published 4.0%; this model uses 4.02% so the worker totals add up. Automation spreading is not the same as jobs disappearing. The rough spending-power score compares payments with an assumed $55,000 yearly earnings amount; it is not measured household income.
What the payment plans leave out: The funding choices only change the assumed price pressure. We do not work out taxes, debt, or money creation in full. A plan that shares productivity gains would still need a real way to collect and pay out the money.
Still to do: Measure AI's effects more reliably, check more phones and browsers, and improve how government funding works. This example does not predict the future or tell you how to invest.
Where the starting numbers come from Starting data and sourcesShows where the starting numbers came from, when they were measured, and how they were used. The starting year stays at 2024 so results can be repeated. The date we downloaded data is not the year it describes. Agencies sometimes revise old numbers; this app keeps a named copy. You can download that same copy and its source notes.
United States · actual 2024 starting numbers. We use a fixed copy of government data so you can repeat the same results. Everything after 2024 is simulated, even years that have already happened. Download date: 2026-10-18. Data version: us-2024-reference-v1.
The economy started at $29.18 trillion at 2024 prices. After removing price changes, that is $23.31 trillion measured using 2017 dollars. One trillion is 1,000 billion.
There were 161.346 million people with jobs. We assume payments go to all 267.455 million resident adults, not only workers.
See each starting number and where it came from
| Starting number | Value | What it measures and when | Official source |
|---|---|---|---|
| Output at each year's prices | 29,184.9 | Billions of US dollars at that year's prices 2024 annual | U.S. Bureau of Economic Analysis 2025-03-27 |
| Output before price changes | 23,305 | Billions of US dollars with price changes removed, using 2017 for comparison 2024 annual | U.S. Bureau of Economic Analysis 2025-03-27 |
| Growth in output before price changes | 2.8 | Percent change in the yearly average 2024 annual | U.S. Bureau of Economic Analysis 2025-03-27 |
| Economy-wide price rises | 2.4 | Yearly percentage rise in economy-wide prices 2024 annual | U.S. Bureau of Economic Analysis 2025-03-27 |
| People working or looking for work | 168,106 | Thousands of people aged 16+, excluding the military and people in institutions 2024 annual average | U.S. Bureau of Labor Statistics 2025-03-05 |
| People with jobs | 161,346 | Thousands of people aged 16+, excluding the military and people in institutions 2024 annual average | U.S. Bureau of Labor Statistics 2025-03-05 |
| Official share of people out of work | 4 | Official yearly percentage, rounded 2024 annual average | U.S. Bureau of Labor Statistics 2025-03-05 |
| Consumer-price score for the previous year | 304.702 | Consumer-price score: the 1982–84 average is 100; regular seasonal changes are not removed 2023 annual average | U.S. Bureau of Labor Statistics 2025-01-15 |
| Consumer-price score for the year | 313.689 | Consumer-price score: the 1982–84 average is 100; regular seasonal changes are not removed 2024 annual average | U.S. Bureau of Labor Statistics 2025-01-15 |
| Growth in output per hour | 2.7 | Yearly percentage growth in output per hour at businesses outside farming 2024 annual | U.S. Bureau of Labor Statistics 2025-03-06 |
| Residents aged 18 and older | 267,454,918 | Number of all residents aged 18 and older 2024-07-01 | U.S. Census Bureau 2026-06 |
Agencies sometimes revise past numbers. This app keeps its named starting copy rather than silently replacing it. The official document titles below are kept as published.
- GDP, fourth quarter and year 2024, third estimate (2025-03-27)
- Regional and State Unemployment, 2024 annual averages, Table 1 (2025-03-05)
- Historical CPI-U, December 2024 release (2025-01-15)
- Productivity and Costs, fourth quarter and annual averages 2024, revised (2025-03-06)
- Vintage 2025 national resident population by age and sex (2026-06)
Which starting numbers have been checked? Checking the starting numbersWe adjust a small leftover growth amount so the starting assumptions match the known 2024 output growth in one check. Matching a known year is not proof the model predicts the future. A separate check trained on two older years did poorly on 2024 and is not used to run the model. We have not measured dependable AI effects or proved which policy is best.
Starting output, jobs, prices, and adult counts come from official figures. Consumer prices rose by 2.95% in the starting data. Prices across the whole economy rose by 2.4%. These measure different things.
The “Today's starting point” story combines 2.7% growth in output per hour with a small 0.1 percentage-point adjustment to match the known 2.8% growth in output for 2024. Choose that story and “No payments” to see the check. Matching one known year does not prove the future results are right.
An older-data test did not predict 2024 well
A separate check used 2022 and 2023 to choose a growth adjustment, then tried to recreate 2024. It gave 5.25% growth compared with the actual 2.8%—a gap of 2.45 percentage points. Simply averaging growth in those earlier years gave 2.7%, a gap of 0.1 points.
Two years is too little evidence for a dependable rule. This check also uses productivity that was known afterward, so it is not a prediction made ahead of time. Its fitted adjustment is not used in the simulator. We show the poor result so the limits are clear.
Which parts are still guesses?
How quickly AI spreads, its effects on jobs and prices, payment rules, the price effects of funding, and the everyday-cost basket are assumptions. The workforce and adult count stay fixed. The same imagined change in price pressure is added to the separate consumer and economy-wide starting rates. We have not measured reliable AI cause-and-effect strengths or worked out a full tax and debt budget.