69
/ 100
Mostly True
United States
The video discusses myths about AI, including its impact on water supply, job displacement, and electricity usage, providing data to counter these claims.
Infact verdict: Mostly True (69/100).
The claims regarding AI's impact on water supply and data center water usage are supported by evidence indicating significant water consumption for cooling purposes. The claim about ChatGPT's water usage per conversation is exaggerated but based on real concerns about AI's environmental impact. The prediction about AI surpassing human radiologists is plausible but remains speculative. The claim about the Mayo Clinic's employment of radiologists contradicts predictions of AI replacing jobs, supported by evidence of increased employment. The historical claim about job evolution is supported by evidence of new job roles emerging since 1940. The prediction about job displacement and creation by 2030 is based on a credible source but remains speculative. The claim about global data center electricity consumption is supported by evidence indicating high energy usage.
How is this score determined? →
▶ Video source
https://youtu.be/vh_y4dth2iU
youtu.be ↗
Individual claims
47
Mixed
Environment
AI is destroying our water supply.
The claim that AI is destroying our water supply is misleading. Evidence shows AI increases water usage for cooling data centers, but companies are addressing this by reusing water and using less water-intensive energy sources. The claim is exaggerated but based on real concerns about AI's environmental impact.
Fact Check Score
None
Fact Check Weight
0
Web Consensus Score
50
Web Consensus Weight
40
Source Quality Score
40
Source Quality Weight
20
Llm Reasoning Score
45
Llm Reasoning Weight
40
Llm Reasoning Score Raw
45
Weighted Total
47
Evidence Summary
3 web sources discuss AI's water usage impact.
43
Mixed
Technology
Every single ChatGPT conversation in the world drinks around one bottle of water.
The claim that each ChatGPT conversation uses a bottle of water is exaggerated. Evidence indicates AI systems consume water for cooling, but the exact amount varies. The claim highlights a real environmental concern but overstates the impact.
Fact Check Score
None
Fact Check Weight
0
Web Consensus Score
45
Web Consensus Weight
40
Source Quality Score
35
Source Quality Weight
20
Llm Reasoning Score
40
Llm Reasoning Weight
40
Llm Reasoning Score Raw
40
Weighted Total
43
Evidence Summary
3 web sources discuss AI water usage.
91
True
Environment
In 2023, every single data center in the US combined evaporated 17.4 billion gallons of water.
The claim is supported by evidence indicating that U.S. data centers used 17.4 billion gallons of water in 2023 for cooling purposes. Multiple sources corroborate this figure, providing a high level of confidence.
Fact Check Score
None
Fact Check Weight
0
Web Consensus Score
90
Web Consensus Weight
40
Source Quality Score
85
Source Quality Weight
20
Llm Reasoning Score
80
Llm Reasoning Weight
40
Llm Reasoning Score Raw
80
Weighted Total
91
Evidence Summary
3 web sources confirm data center water usage.
94
True
Environment
The United States uses 322 billion gallons of water every single day.
The claim is supported by evidence from authoritative sources like the U.S. Geological Survey, confirming that the U.S. uses 322 billion gallons of water daily. This figure is well-documented and reliable.
Fact Check Score
None
Fact Check Weight
0
Web Consensus Score
95
Web Consensus Weight
40
Source Quality Score
90
Source Quality Weight
20
Llm Reasoning Score
85
Llm Reasoning Weight
40
Llm Reasoning Score Raw
85
Weighted Total
94
Evidence Summary
3 authoritative sources confirm U.S. water usage.
91
True
Environment
Golf courses in the United States consume around 2 billion gallons of water every single day.
The claim is supported by evidence indicating that U.S. golf courses consume around 2 billion gallons of water daily. This figure is corroborated by multiple sources, providing a high level of confidence.
Fact Check Score
None
Fact Check Weight
0
Web Consensus Score
90
Web Consensus Weight
40
Source Quality Score
85
Source Quality Weight
20
Llm Reasoning Score
80
Llm Reasoning Weight
40
Llm Reasoning Score Raw
80
Weighted Total
91
Evidence Summary
3 web sources confirm golf course water usage.
27
Mostly False
Health
AI would read medical scans better than any human alive within 5 years.
This claim is a prediction about future AI capabilities. While studies suggest AI's potential to surpass human radiologists, it remains speculative and cannot be verified at this time.
Fact Check Score
None
Fact Check Weight
0
Web Consensus Score
None
Web Consensus Weight
40
Source Quality Score
None
Source Quality Weight
20
Llm Reasoning Score
50
Llm Reasoning Weight
40
Llm Reasoning Score Raw
50
Weighted Total
27
Evidence Summary
Prediction about AI surpassing human radiologists.
91
True
Health
The Mayo Clinic employs 55% more radiologists than it did when Hinton spoke.
The claim is supported by evidence indicating that the Mayo Clinic employs 55% more radiologists than in 2016, contradicting predictions of AI replacing these jobs. This is corroborated by multiple reliable sources.
Fact Check Score
None
Fact Check Weight
0
Web Consensus Score
90
Web Consensus Weight
40
Source Quality Score
85
Source Quality Weight
20
Llm Reasoning Score
80
Llm Reasoning Weight
40
Llm Reasoning Score Raw
80
Weighted Total
91
Evidence Summary
3 sources confirm increased employment of radiologists.
88
True
Economics
60% of the work Americans do today did not exist in 1940.
The claim is supported by evidence indicating that 60% of today's jobs did not exist in 1940, highlighting the evolution of the workforce. This is corroborated by multiple sources, providing a high level of confidence.
Fact Check Score
None
Fact Check Weight
0
Web Consensus Score
85
Web Consensus Weight
40
Source Quality Score
80
Source Quality Weight
20
Llm Reasoning Score
75
Llm Reasoning Weight
40
Llm Reasoning Score Raw
75
Weighted Total
88
Evidence Summary
3 sources confirm job evolution since 1940.
27
Mostly False
Economics
The World Economic Forum predicts 92 million jobs will be displaced and over 170 million created by 2030.
This claim is a prediction based on a credible source, the World Economic Forum. While the prediction is plausible, it remains speculative and cannot be verified at this time.
Fact Check Score
None
Fact Check Weight
0
Web Consensus Score
None
Web Consensus Weight
40
Source Quality Score
None
Source Quality Weight
20
Llm Reasoning Score
50
Llm Reasoning Weight
40
Llm Reasoning Score Raw
50
Weighted Total
27
Evidence Summary
Prediction about job displacement and creation by 2030.
91
True
Environment
All data centers on Earth combined used around 400 terawatt hours in 2024.
The claim is supported by evidence indicating that global data centers consumed around 415 terawatt hours of electricity in 2024. This figure is corroborated by multiple sources, providing a high level of confidence.
Fact Check Score
None
Fact Check Weight
0
Web Consensus Score
90
Web Consensus Weight
40
Source Quality Score
85
Source Quality Weight
20
Llm Reasoning Score
80
Llm Reasoning Weight
40
Llm Reasoning Score Raw
80
Weighted Total
91
Evidence Summary
3 sources confirm global data center electricity consumption.
