How Does Brain Efficiency Relate to Intelligence?

TL;DR

Higher intelligence-test scores were associated with lower glucose metabolism in active brain regions in Richard Haier's first PET study, leading to the brain efficiency hypothesis. Haier stresses that this interpretation remains testable rather than settled: no brain story is simple, no single study is definitive, and reliable conclusions emerge only after years of comparing conflicting evidence.

Transcript

I'm here with Professor Richard Haier, who has studied the neuroscience of intelligence for 40 years or so. Is that correct? 40 years or more. It's not quite clear to me, but a long time. Right. So what are you working on these days? Well, I just, a couple of months ago, finished the manuscript for a new textbook called the Science of Human Intelli... Read More

Key Insights

• The brain efficiency hypothesis arose from an inverse correlation between glucose metabolism and intelligence-test scores: participants whose brains used more energy in task-related regions tended to receive lower scores on the nonverbal abstract reasoning test.
• Scientific theories of intelligence are provisional explanations derived from empirical observations, not declarations of certainty. Haier describes his theories as best guesses formulated to produce hypotheses that other researchers can test, challenge, refine, or falsify.
• No single neuroscience study is definitive because brain processes are complex and findings can conflict. Haier argues that researchers need many years of accumulated evidence before they can judge the overall support for a claim about intelligence and the brain.
• Evidence concerning brain efficiency is mixed rather than conclusive. A researcher reviewing the field judged the supporting evidence to be weak, and Haier welcomed that assessment because testing the hypothesis had still generated discoveries and new questions.
• PET imaging works by injecting a radioactive tracer while a person performs a cognitive task. Brain regions using more glucose absorb more tracer and produce more radioactivity, allowing researchers to construct an image of areas consuming the most energy.
• Early intelligence imaging research faced severe practical limits. Each PET scan cost $2,500, the scanner cost more than a million dollars, a cyclotron was required to produce short-lived radioisotopes, and the first reported study included only eight participants.
• Modern brain research has access to dramatically larger datasets than the earliest PET studies. Haier contrasts an original eight-person sample with public databases containing images from a thousand people that researchers can access without paying for each individual scan.
• Technological progress changes which intelligence questions scientists can investigate. Haier notes that fMRI became broadly available years after his early PET work and urges younger researchers to anticipate the technologies and research questions that may emerge over the next 35 years.
• More videos from this channel:
  • How relativity explains consciousness with Nir Lahav
  • How Does Space-Time Emerge From Quantum Entanglement?
  • How Will AI Transform Mathematical Research?
• More from this channel: Curt Jaimungal

Questions & Answers

Q: What is the brain efficiency hypothesis?

The brain efficiency hypothesis proposes that stronger performance on an intelligence test may depend on how efficiently the brain works rather than how much energy it consumes. It originated after Haier's first PET study found inverse correlations between glucose metabolic activity in task-related brain regions and scores on nonverbal abstract reasoning problems. Participants showing greater metabolic activity tended to earn lower scores.

Q: How does PET imaging measure brain activity?

Positron emission tomography measures activity by injecting a radioactive tracer while a participant performs a cognitive task. The tracer marks brain regions that take up the most glucose and therefore use the most energy. Regions working harder absorb more tracer and emit more radioactivity, which enables researchers to create an image showing the distribution of metabolic activity.

Q: What did Haier's first PET intelligence study find?

Haier's first PET intelligence study examined eight people solving nonverbal abstract reasoning problems. Some brain regions were more active during problem solving than during control conditions. The unexpected result appeared when metabolic activity was correlated with test performance: greater glucose metabolism in the active regions was associated with lower scores, inspiring the brain efficiency interpretation.

Q: Why is one brain-imaging study not definitive?

A single brain-imaging study cannot settle a question because brain explanations are complex, individual results may conflict, and early studies can have serious practical limitations. Haier's first experiment, for example, included only eight participants. He argues that researchers must compare many studies over years before the accumulated weight of evidence can support, revise, or reject a hypothesis.

Q: How strong is the evidence for brain efficiency?

The interview presents the evidence as mixed and still open to evaluation. Haier recounts a research review that found support for the brain efficiency hypothesis was not especially strong. He regarded that result positively because the proposal was always meant to be tested, and subsequent research produced useful findings even when it failed to confirm the original formulation consistently.

Q: What were the challenges of early PET research?

Early PET research required costly and scarce infrastructure. Haier says each scan cost $2,500, a scanner cost more than a million dollars, and the short-lived radioisotopes required the researchers to obtain their own cyclotron. There was no ready-made image-analysis software or accessible image database, so even collecting a sample of eight participants represented pioneering work.

Q: How should scientific theories of intelligence be evaluated?

Scientific theories of intelligence should be treated as testable explanations derived from observations. Haier encourages readers to examine the studies, locate weak links or loopholes, and design research that can correct or falsify the proposed account. The goal is not to defend a theory regardless of evidence, but to revise hypotheses as new, inconsistent, or contradictory findings accumulate.

Q: How has brain-imaging research changed since the 1980s?

Brain-imaging research has moved from rare, expensive PET facilities and tiny samples toward broadly available technologies and large public datasets. Haier recalls working before standard analysis software or image databases existed, when eight scans were significant. He contrasts that environment with researchers being able to access brain images from a thousand people without paying for each scan.

Summary & Key Takeaways

• Richard Haier describes a scientific career focused on the neuroscience of intelligence and several books intended for different audiences. His introductory writing reviews the research literature while openly presenting his interpretations as provisional. He invites readers to identify weak links, design better studies, and attempt to falsify proposed explanations through empirical research.

• Haier offers three guiding principles for neuroscience: brain explanations are never simple, individual studies are not definitive, and conflicting findings can require many years to resolve. He treats theories as generators of testable hypotheses, arguing that even evidence against an original proposal can produce valuable discoveries and motivate stronger theories based on newer data.

• Early PET research measured glucose metabolism while eight participants solved nonverbal abstract reasoning problems. Contrary to expectations, higher scores correlated with less metabolic activity in relevant brain areas. Haier and colleagues interpreted this inverse relationship as possible brain efficiency, although later reviews found inconsistent and contradictory evidence that prevented a simple, definitive conclusion.