The MIT Siegel Family Quest [SQI] is for AI, not Human Intelligence

By David Stephen

If an organization [academic, nonprofit or industry] say they want to solve or understand how human intelligence works, and in their statement, begin to mention [the improvement of] artificial intelligence, you should immediately know that they are just another AI lab, to make AI or robots better.

There is no other focus to learn about how human intelligence works than the brain. And in the brain, to understand human intelligence is to leap neurons and totalize on electrical and chemical signals, as the drivers of human intelligence.

MIT Siegel Family Quest [SQI] is for AI

What is referred to as neurons firing, excited or active are the voyages of electrical and chemical signals along them. When they don’t [fire] electrical and chemical signals make determinations, as well.

So, whatever intelligence is, is directly mechanized, at least conceptually, by electrical and chemical signals.

Though it is true that they are everywhere in the brain, and found in other places too in the central and peripheral nervous systems, what makes their effort for intelligence different from their effort for say, memory, emotions, feelings and interoception? This is a working question to make progress on human intelligence.

An organization cannot claim to work on human intelligence without having basic questions or have outstanding postulates on the direction to go. An organization cannot also claim to want to work on human intelligence without formulating a robust definition for human intelligence, which can be applied universally.

An organization that also wants to work on human intelligence must, at least, proscribe any connection of their research to benefit or improve AI, for as long as possible. Why? Because, by making AI an option for improvement, progress in their work would be adapted to AI and robotics, even when they cannot answer fundamental questions about human intelligence in the brain.

Understanding and exploring to improve human intelligence has nothing to do with artificial intelligence. Artificial intelligence research is already oversubscribed and in intense efforts for advancement across industry and academic labs.

There is nothing comparable for human intelligence. There is currently no human intelligence research lab anywhere on earth, even at the ‘MIT’.

Now, a team that has human intelligence as a mission must make it a strict pursuit, not confused with artificial intelligence, which would eventually benefit artificial intelligence.

MIT

There is a new [January 14, 2026] announcement on MIT News, At MIT, a continued commitment to understanding intelligence, stating that, “The MIT Siegel Family Quest for Intelligence (SQI), a research unit in the MIT Schwarzman College of Computing, brings together researchers from across MIT who combine their diverse expertise to understand intelligence through tightly coupled scientific inquiry and rigorous engineering. These researchers engage in collaborative efforts spanning science, engineering, the humanities, and more.”

“SQI seeks to comprehend how brains produce intelligence and how it can be replicated in artificial systems to address real-world problems that exceed the capabilities of current artificial intelligence technologies.”

“In SQI, we are studying intelligence scientifically and generically, in the hope that by studying neuroscience and behavior in humans and animals, and also studying what we can build as intelligent engineering artifacts, we’ll be able to understand the fundamental underlying principles of intelligence,”

“We in SQI believe that understanding human intelligence is one of the greatest open questions in science — right up there with the origin of the universe and our place in it, and the origin of life. The question of human intelligence has two parts: how it works, and where it comes from. If we understand those, we will see payoffs well beyond our current imaginings.”

“There has been remarkable progress in AI over the past decade, but I believe the next decade will bring even greater advances in our understanding of human intelligence — advances that will reshape what we call AI. By supporting us, David Siegel, the Siegel Family Endowment, and our other donors are demonstrating their confidence in our approach.”

“We’re part of the Schwarzman College of Computing, at the nexus between the people interested in biology and various forms of intelligence and the people interested in AI. We’re working with partners at other universities, in nonprofits, and in industry — we can’t do it alone.”

“Fundamentally, we’re not an AI effort. We’re a human intelligence effort using the tools of engineering.”

“We need to focus on problems that mirror natural and artificial intelligence — making sure that we are evaluating new models on tasks that mirror what humans and other natural intelligence can do.”

Evaluating Human Intelligence Research

How do you evaluate an effort in human intelligence research, especially if the announcement is from a so-called prestigious institution?

They can say they are ‘not an AI effort’ or ‘don’t care whether there are commercial applications for this quest’ or whatever, but how do you say so much in an announcement and not mention a directional hypothesis about the possible components of intelligence in the brain and how they might be working to result in intelligence?

Also, study human intelligence for what? They mentioned engineering a lot, which is a code word for AI. A better goal is to explore human intelligence for problem-solving.

This means that if humans, wherever, would solve problems in communities, what can be understood about how human intelligence works to make the objective more likely?

Simply, embracing AI in their research, no matter how effective will be AI-centered human intelligence. It would, at minimum, require facilities to use or access AI. Real and beneficial progress will be the ability for humans to adventure valuable problem-solving, without necessarily having facilities, or say AI.

So, if someone wants to solve a certain problem, what should be in memory, when should it be thought about, what factors must be considered, what might be seen or heard that might ensure that relays in the brain produce excellent results?

Seeking to know how human intelligence works to improve problem-solving is better than the vague engineering speak or robotics [euphemism: embodied intelligence] hopes.

Also, the primary focus to study human intelligence is the brain, or postulates about how electrical and chemical signals drive it.

At least, that is where to make progress, at that fundamental science level, not engineering or humanities, yet. Lumping a lot of fields, when the main goal, the brain seems secondhand, is indicative of a mission astray.

The MIT Siegel Family Quest for Intelligence [SQI] is not the hope of humanity to solve human intelligence, not in this decade or in this century. Even just studying electrical signals in the brain alone, could answer more questions towards intelligence than this whole heap of soured methodology.

MIT Quest YouTube Channel is all AI

Everything on SQI’s YouTube channel promotes AI, artificial or robotics. Even their talks on brain, it all serves artificial in conclusion or points in the direction. Showing that their work is optimized for AI.

They have: “The Next Horizon: Embodied Intelligence Mission.

The Next Horizon: Perceptual Intelligence Mission.

The Next Horizon: Language & Thought Mission.

The Next Horizon: Development of Intelligent Minds.

Foundation Models of Human Cognition.

From Thought to Movement: Helping Paralyzed People with Brain-Machine Interfaces.

Cognitive Tools for Making the Invisible Visible.

Discussion: Language and The Development of Thought – Matter of Minds

Insights into AI algorithms drawn from hippocampal function.”

Their mission, they stated is around ‘Perceptual Intelligence, embodied intelligence, development of intelligence, language, social intelligence.’

They are looking for ‘machine executable models’. Or, they want ‘to understand intelligence in engineering terms’. Or, they are working on ‘integrated computational models of past sensory intelligence’.

All of these are distant from a theory to result in a simple mind display of what happens when an individual does an intelligent task.

Human intelligence is defined as the use of memory for expected, desired or advantageous outcomes. There are two main types, operational intelligence and improvement intelligence.

Intelligence is not found in just some areas of the brain. Also, some of the same attributes that electrical and chemical signals use in other functions apply to intelligence.

These are better than whatever MIT Quest has done for years or what SQI would do.

MIT Neuroscience is Primitive

At MIT neuroscience, they believe that thought and language are different because neuroimaging lights up separate areas of the brain, for thought processes and language processes, [according to one of their talks].

The problem with this assumption is how labels have driven their neuroscience and others in the wrong direction.

First, the same components [electrical and chemical signals] are involved in both processes. So, why do they specify one and differentiate others?

Secondly, aside that some areas overlap in activity with thought and language, it is almost impossible to not question that thinking about a door and speaking, listening, writing or reading do not involve extended configurations [even in different areas] or at least a collective assembly of signals.

Simply, if one area is silent in one process and active in the other, even though the functions [say door are similar], is it guaranteed that they are actually different? The sight of a door, the sound of a door and door in a language use case? How is door, as an information, organized in the brain? When door is seen or heard, what does the brain use and how do scenarios make the source different? Also, assuming the brain is storing respective specifications, how efficient would that be?

It is theorized that door is a thick set of electrical and chemical signals. So, whatever is common between two or more doors are stored in that thick set. Now, some other thick sets overlap with areas of the thick set of a door, where similarities exist. It is possible that a part of a thick set exists in an area and the rest in other areas. However, thick sets have special sequences [new or old], for transport between them. Thin sets hold the most unique information about a specific function that does not exist elsewhere. So, a particular type of door. This thin set too could be somewhere while connecting the rest of the information elsewhere with the thick set.

Principally, there is a way [sets of] electrical and chemical signals organize information. There is also a way that they transport. This is the actual quest, then to explain how neuroimaging matches with that. Not to simply assume that differences in activity area means differences of functional labels.

To make progress in brain science, the first thing is to throw away all labels. Then look at the implicated components within the cranium and start to directly understand or propound how they might be working.

Then circle back, maybe, to labels. But to start and stick with labels is just off-track.

Intelligence too is a label. If memory is a destination and intelligence is the use of memory, it means that intelligence is mostly linked with relays.

So, what relays define intelligence? These are answers to seek for progress. SQI is no different from any other AI lab.

They are not for human intelligence. They were not for human intelligence. They will not be for human intelligence.

Unfortunately for David Siegel, he saw a facade, picked a circus, and flushed his resources.

David Stephen currently does research in conceptual brain science with focus on the electrical and chemical configurators for how they mechanize the human mind with implications for mental health, disorders, neurotechnology, consciousness, learning, artificial intelligence and nurture. He was a visiting scholar in medical entomology at the University of Illinois at Urbana Champaign, IL. He did computer vision research at Rovira i Virgili University, Tarragona.

See more breaking stories here.