Pass Talk and Writing: Fit AI Let People Dey Exchange Thoughts at High Speed?

One scientific study of artificial intelligence, brain-computer interfaces and direct thought exchange between people

Scientific review article and proposed research framework | November 2026

Abstract

Today modern computers dey exchange information at speeds wey hard to imagine.

One ordinary Gigabit Ethernet connection fit transfer data at one billion bits (1,000,000,000 bits) per second. Advanced computing and data-centre technologies dey operate many times faster pass dat one.

But di human situation dey different kpatakpata.

Even today, people still dey mostly use talk, writing, gestures and body movements to carry dia thoughts, ideas, opinions and feelings give other people.

Even though plenty-plenty neural information processing dey happen inside di human brain, our capacity to transfer dat information to another person in a meaningful way dey very limited wen you compare am.

Dis situation dey raise one important scientific question.

Using artificial intelligence and neural technologies, fit people exchange dia thoughts and ideas directly, accurately and at high speed, without di help of talk or writing?

Dis article dey examine di speed of human information transfer, technologies wey dey interpret di brain's neural signals, brain-computer interfaces, brain-to-brain communication and di scientific progress wey don reach November 2026.

E also dey propose one new research framework for di high-speed exchange of human thoughts and intentions with di help of artificial intelligence.

Di fundamental aim of dis research no be just to transfer more digital data.

Im central purpose na to explore di scientific feasibility of one new communication method wey go carry di true meaning of wetin one person wan express give other people, in less time and with better accuracy.

1. Introduction: Di Speed Barriers for Human Communication

Make you consider two computers wey connect to one modern computer network.

Dem fit convert very large amounts of information into electrical signals and exchange dem inside just few milliseconds.

Now make you consider two people wey dey exchange one complex idea.

Di first person must form di idea inside im mind.

Den e must convert dat idea into language.

E must choose di correct words.

E must arrange dose words into proper sentence.

Na only den e fit talk or write.

Di same way, di second person must hear or read dose words. Den e must understand dia meaning and reason wetin di first person really wan talk.

Every stage of dis process dey take time.

For some occasions misunderstanding fit happen. Di idea go con need explanation again.

Di important problem here no be say di human brain dey work slow.

Rather, na say big gap dey between di neural information processing wey dey happen inside di human brain and di speed wey dat information fit take reach other people.

Research by Coupé and im colleagues, wey dem publish for di journal Science Advances for 2019, study 17 languages and estimate say di information-transfer rate of human spoken languages na about 39 bits per second for average [1].

For another important study, Zheng and Meister estimate di rate of human behavioural information processing at roughly 10 bits per second. Dat study publish online for 2024 and appear for di journal Neuron for 2025 [2].

Dis two studies dey measure different kinds of information processing. So di two numbers no fit replace each other.

And dem no be di final measurement of di human brain's total neural processing capacity.

All di same, dem dey show say di information rate of conscious human actions and communication dey far lower pass di data-transfer rate of many electronic systems.

Dis difference na im form di foundation of our new research.

2. Di Data-Transfer Speed of Computers and di Communication Speed of Humans

Computers dey convert information into digital form and dey exchange am according to defined communication protocols.

For example:

• 1 gigabit per second (1 Gbps) = 1,000,000,000 bits per second.

• 10 Gbps = 10 billion bits per second.

• 100 Gbps = 100 billion bits per second.

• 800 Gbps = 800 billion bits per second.

Dis figures dey describe di information-transfer capacity of one computer network.

But dem no talk anything about how fast di computer dey understand di true meaning of di information wey dey send.

Human communication dey different kpatakpata.

For exchanges between people, wetin matter pass no be only di amount of information but im meaning.

For example, one person fit express one very complex decision clearly with just few words.

Another person fit need several paragraphs to explain di same idea.

So great scientific care dey needed wen you dey compare di data-transfer speed of computers with di communication speed of humans.

One computer fit transfer one billion bits per second.

But dat one no mean say e understand di meaning of dat information.

At di same time, one person fit carry one deep idea give another person with only few words.

So di aim of our research no be to turn di human brain into one gigabit computer network.

Rather, our aim na to increase di speed wey meaningful thoughts and ideas fit take exchange between people in a way wey dem go understand am correctly.

Dis na di most important scientific foundation of dis research.

3. Di Progress Wey Science Don Make So Far

3.1 Brain-Computer Interfaces

Brain-computer interface, or BCI for short, na technology wey dey create connection between di neural activity of di human brain and one external computer system.

Using dis technology, signals wey dem get from di brain fit dey analysed by computer.

Two main approaches dey for recording brain activity.

First, methods like EEG wey dey record di brain's electrical activity without surgery.

Second, methods wey dey use electrodes wey dem implant inside di brain through surgery.

Artificial intelligence models dey analyse dis neural signals and dey try identify di action wey di person intend to do, di word wey e wan express, or one communication command.

Research by Willett and im colleagues, wey publish for di journal Nature for August 2023, na important advance for dis area.

Dat work create one technology wey dey convert into text di neural activity wey dey happen wen one person dey try to talk.

Di participant fit communicate at di rate of about 62 words per minute.

Di word error rate na 9.1% with restricted vocabulary of 50 words and 23.8% with large vocabulary of 125,000 words [3].

Dis research prove say brain signals wey dey produced wen one person dey try to talk fit dey used to communicate relatively fast.

But dis system no be technology wey fit freely read every thought wey dey arise for one person's mind.

E identify di neural activity wey connect with di attempt to talk.

3.2 Generating Speech Directly from Brain Signals

More notable advances for dis field happen for 2025.

Littlejohn and im colleagues create one technology wey dey process signals from di human brain in 80-millisecond intervals and dey convert dem into speech with di help of artificial intelligence [4].

For another study, Wairagkar and im colleagues describe one method wey dey use di brain's neural signals to produce expressive synthetic speech in real time.

Di participant for dat study fit also control, to some extent, expressive elements like voice intonation [5].

Dis na very important advance.

Human communication no be just di exchange of words.

Di speed wey we take dey talk, di tone of voice, emotional expression and di way we dey stress words all dey reveal di true meaning of one idea.

Dis studies demonstrate di feasibility of reconstructing natural human speech from neural signals.

All di same, dem still be experimental technologies wey dem dey use for specific research settings.

3.3 Long-Term Practical Use: One Landmark Study of 2026

One study wey publish for di journal Nature Medicine for June 2026 show important advance for di practical use of brain-computer interface technology.

For dis study by Card and im colleagues, one person with severe motor paralysis communicate independently for long period with di help of one implanted interface.

For nearly two years, e use di technology for im house for more pass 3,800 hours.

Through am e communicate about 1.96 million words and 183,060 sentences.

Di average communication speed na about 56 words per minute.

For controlled speech tests, word accuracy pass 99%. But for everyday natural conversations dis accuracy no be di same at all times [6].

Dis difference dey very important.

Na one achievement for technology to work well for laboratory.

Na another achievement for di same technology to work reliably for long period inside people's daily lives.

For our high-speed human thought-exchange research, both of dis aspects must also dey proven separately.

3.4 Fit Information Dey Exchanged Directly Between Human Brains?

Some early successes don also happen for di scientific world about brain-to-brain communication.

One study wey dem call BrainNet publish for 2019.

For dat study, three people work together on one simple game using brain signals.

Di decisions of di first two participants na EEG technology record dem.

Dat information send go di third participant as defined signals through one brain-stimulation method wey dem dey call Transcranial Magnetic Stimulation (TMS).

Five groups take part for dis experiment.

Dia average performance accuracy record as 81.25% [7].

Dis study demonstrate di feasibility of transferring certain information between human brains through neural interfaces.

3.5 Detecting Speech Wey Person Dey Imagine for Mind

One study by Kunz and im colleagues wey publish for di journal Cell for August 2025 show say di neural activity wey dey happen wen one person dey imagine say e dey talk inside im mind fit dey identified partially.

Another important aspect of dis research na mental privacy.

Di risks wey one person's private inner speech, wey e no wan share with other people, fit accidentally reveal through technology na dem also examine am.

In addition, safety methods to distinguish between thoughts wey dem deliberately create for communication and private mental activity na dem test am [8].

Another study, wey publish for di journal Nature Neuroscience for September 2026, examine some common structures inside di neural activity of imagined speech and overt speech [9].

Dis findings fit create opportunities for di future for people to communicate certain information without even attempting to talk.

But no scientific evidence dey to say current technologies fit accurately understand every thought wey dey arise for any person's mind, without prior training or cooperation.

4. Di Proposed New Research Concept

High-Speed Human Thought Exchange with di Help of Artificial Intelligence

Artificial Intelligence-Assisted High-Speed Thought Communication

Di central aim of dis research:

To explore di feasibility of one technology wey dey identify, from di brain's neural signals, di thoughts and ideas wey people voluntarily wan communicate, dey convert dem into efficient digital form, dey transfer dem to another person, and dey enable dat person to understand di true meaning of di idea accurately.

Dis research dey bring several scientific fields together.

Brain-Computer Interfaces: na dem dey use record and analyse signals wey dem obtain from di human brain.

Artificial Intelligence: na dem dey use am for tasks like detecting patterns inside neural signals, language generation, understanding ideas and correcting errors.

Information Theory: e dey help assess mathematically how efficiently and accurately information fit dey transferred.

Secure Digital Networking: e dey help send di generated information safely to another computer system.

Brain stimulation and neural technologies: dem fit help explore di future possibility of delivering specific information directly to di recipient's neural system. Extensive medical and scientific testing dey essential for dis one.

Di important opportunity for dis research fit dey for creating one method wey dey carry di meaning of wetin one person wan express, without needing to generate and transfer every single word wey e wan talk.

For example, make we suppose say one person dey think like dis:

"I agree with dis plan. But I go like to postpone di implementation reach next week."

One future technology fit try identify di three important elements inside dis thought.

First, agreement with di plan.

Second, reluctance to implement am immediately.

Third, request to postpone am to next week.

Artificial intelligence fit identify dis basic meanings and construct one sentence wey di other person fit understand.

But one important challenge dey here.

E no dey enough say di sentence wey artificial intelligence generate dey elegant for language.

E must express, without change, di idea wey di first person really wan convey.

Dis na di importance of di concept wey dem dey call Semantic Communication.

5. Di Proposed Technical Framework

Di initial structure for implementing dis research fit divide into six main stages.

Stage 1: Recording Brain Signals

One person decide, with im full consent, to convey information to another person.

At dat moment, di brain activity wey connect with di word, command or idea wey e deliberately dey try to express go dey recorded with di help of one suitable neural interface.

For early-stage research, non-surgical methods like EEG and already published research data fit dey used.

Stage 2: Identifying di Thought with Artificial Intelligence

Di recorded brain signals go dey analysed by machine-learning models.

Di aim na to identify di information wey di person deliberately try to convey.

Research suppose begin by identifying small number of words, commands or ideas.

Na only after di accuracy of dat basic technology don dey confirmed work suppose move on to understanding complex thoughts.

Stage 3: Converting di Thought into Meaning-Based Data

Di identified information go dey converted, with di help of artificial intelligence, into one structured digital form.

Dis one fit include di thought's main ideas, di relationships between dem, conditions and di uncertainty of di estimate.

Dis one dey make am possible to represent di basic meaning of di information inside one computer system even before complete sentence dey generated.

Stage 4: Secure High-Speed Information Transfer

Di generated data go dey sent to another system through one secure, encrypted digital communication channel.

At dis stage, di speed of di computer network no need to be one major barrier.

Today's computer networks don already dey capable of transferring information at very high speed.

Instead, accurately understanding brain signals, generating di correct idea and reducing di time wey di whole process dey take fit be di more difficult problems.

Stage 5: Reconstructing di Message for di Recipient

Artificial intelligence dey convert di received information into one form wey another person fit understand.

E fit be text, speech or another suitable form of communication.

Di recipient must dey enabled to understand di meaning of di information, to ask for clarification if e dey in doubt, and to know di uncertainty inside di artificial intelligence's estimate.

Stage 6: Confirming Say di Message Don Dey Understood

E must dey confirmed say di recipient don understand di idea correctly.

Where e dey needed, di facility to request clarification or correction must also dey provided.

If direct brain-to-brain communication go dey created for di future, one separate technology wey fit deliver information to di recipient's neural system must also dey developed.

One important distinction must dey noted here.

Converting one person's brain signal into text by computer and make another person read am for screen na brain-computer-human communication.

Dem no fit call am complete direct brain-to-brain communication.

Dis distinction must dey maintained clearly at every stage of di research.

6. Di Central Scientific Hypothesis of Dis Research

Di central hypothesis for dis research dey proposed as follows.

By identifying, with di help of artificial intelligence, di neural information wey people deliberately wan communicate, and converting im meaning into one efficient digital form, certain ideas fit dey conveyed to other people in less time pass with talk or writing, while maintaining acceptable accuracy.

Dis hypothesis fit dey tested scientifically.

For dis one, di same kind of information fit dey exchanged through ordinary talk, writing and neural interfaces wey connect to artificial intelligence.

Wherever e dey possible, di same participants and di same information suppose dey used to compare di results.

Di following measurements dey important for dis study.

1. End-to-End Latency

Di time from wen one person begin to convey information until another person correctly receive am.

2. Effective Information-Transfer Rate

One measure of how much information correctly transfer per second, after accounting for errors and corrections.

3. Semantic Fidelity

Di similarity of meaning between di idea wey di first person wan convey and di idea wey di second person understand.

4. Error Rate

Di amount of words, ideas, relationships and conclusions wey dem identify incorrectly.

5. Receiver Comprehension

One measure of whether di person wey receive di information correctly understand im true meaning.

6. User Effort

Di attention, mental strain or fatigue, training and ease of use wey dey required to use di technology.

7. Privacy Protection

Monitoring whether thoughts or information wey di user no authorise accidentally dey disclosed.

Dem no fit say one technology don accelerate communication between people merely because im network connection dey fast.

Di same way, di fact say artificial intelligence dey generate one sentence very fast no be sufficient evidence.

Di complete human-to-human communication process must dey measured.

7. Di Proposed Experimental Research Plan

Dis research fit dey carried out in three main stages.

Stage One: Identifying Human Intentions

Initially, dem must investigate whether few specific ideas wey people deliberately wan convey fit dey identified from brain signals.

Already published EEG datasets, or new studies wey dem conduct with proper consent and permission, fit dey used.

For example, participants fit attempt to convey one of di following ideas.

• I agree.

• I no agree.

• I need help.

• I need further explanation.

Machine-learning models fit dey used to examine whether differences dey appear inside di brain signals wey connect with dis ideas.

Di model's performance must den dey evaluated using data wey dem no use for training, different test sessions and independent participants.

Di important question for dis stage:

E dey possible to identify reliably, from brain signals, di idea wey one person wan convey?

Stage Two: Meaning Transfer with di Help of Artificial Intelligence

One attempt fit den dey made to convert di information wey dem identify for stage one into complete ideas with di help of artificial intelligence.

Three aspects must dey compared here.

First, di idea wey di sender really wan convey.

Second, di idea wey di artificial intelligence reconstruct.

Third, di meaning wey di person wey receive dat information understand.

How accurately dis three correspond must dey examined through mathematical and statistical methods.

Most importantly, wen one idea no dey clear to di artificial intelligence, e must express im uncertainty instead of generating incorrect information.

Di performance of dis method must dey compared with ordinary typing, talk and other communication methods.

Stage Three: Communication Between People

For di third stage, information-transfer experiments fit dey conducted between two people wey dey participate with full consent.

Di first person deliberately dey attempt to convey information through one brain signal.

Artificial intelligence dey analyse dat signal, dey identify di relevant information and dey send am securely.

Di second person dey receive and understand dat information.

Di speed, accuracy and reliability of di complete communication process den dey measured.

Delivering information directly to di recipient's brain go be one much more advanced and separate research stage.

Such attempts dey require neural expertise, regulatory approvals, medical safety and strict ethical controls.

So priority suppose dey given, for early-stage research, to proving whether artificial intelligence fit improve di performance of communication between people.

8. Scientific Challenges and Current Limitations

Several important challenges dey for dis research.

First, no be all human brains dey represent words and ideas with di same neural patterns.

So each person fit need individual training or model adjustment.

Second, signals wey dem record for non-surgical methods like EEG fit contain high levels of noise and interference.

Compared with electrodes wey dem implant inside di brain, di ability to detect certain kinds of neural activity in fine detail fit also dey lower.

Third, attempting to talk, imagining speech for mind, intending to express one idea and thinking privately to oneself no be di same kind of neural activity.

Dem must dey studied separately.

Fourth, large language models fit sometimes generate sentences wey no dey true but wey dey appear plausible, based on di information wey dey available to dem.

For human thought-exchange technology, such errors fit become one very large problem.

Telling somebody say one person think something wey e no think fit create serious misunderstandings.

Fifth, e dey very important say di results of scientific studies fit dey reproduced by other researchers.

One study wey publish for di Journal of Neural Engineering for 2026 show say plenty problems dey for reproducing di results of studies wey dey identify imagined speech.

Among di speech-imagery datasets wey dem examine for dat study, only 36% of participants achieve classification accuracy wey cross di appropriate statistical significance threshold [10].

Dis one demonstrate how important independent verification, transparent research methods and reliable testing dey for dis field.

Finally, di speed of information transfer no be di only challenge.

Human attention, memory, comprehension and thought processes also dey play important roles.

Removing di need to talk no dey, by itself, make di time wey dey needed to form or understand one complex idea disappear entirely.

Based on current scientific evidence, dem no fit say every kind of human thought fit dey transferred directly at gigabit speed.

Treating digital data-transfer speed and di transfer speed of meaningful human thoughts as di same thing na scientifically incorrect approach.

9. Privacy, Information Security and Human Rights

Technologies wey dey interpret di neural signals of di human brain dey also create very important ethical responsibilities.

For ordinary computer systems, di information wey dey processed na information wey people enter by demself.

But neural interfaces fit also record signals wey connect with one person's private mental activity wey e no wan share with other people.

So di most important basic principle of our research must be di following.

Only di information wey one person, with im full consent, deliberately wan share suppose dey identified and sent to other people.

Several safety arrangements dey necessary to put dis principle into practice.

Di user must get explicit control to initiate communication.

Wherever e dey possible, neural data suppose dey processed on di user's own device.

Information must dey transferred in encrypted form.

Strong security systems must dey created to prevent unauthorised persons from accessing brain signals.

Storing unnecessary neural data for long periods must dey avoided.

Di user must get di facility to stop di transfer of information immediately at any time.

In addition, e must dey made clear say di ideas wey artificial intelligence dey generate no be one-hundred-percent certain records of di thoughts inside one person's mind.

Dem be estimates wey dem generate on di basis of neural signals.

For November 2025, UNESCO adopt one international recommendation on di ethical use of neurotechnologies [11].

Dat recommendation dey give importance to mental privacy, human dignity and di responsible use of neural technologies.

For our research too, privacy and ethical safeguards must dey waka together with technological development.

10. Future Applications of Dis Technology

If dis research succeed, im early benefits fit reach people wey dey face severe difficulties with talk and physical movement.

In particular, brain-computer interfaces fit create new opportunities for those wey no fit communicate because of motor paralysis and certain neurological conditions.

Such technologies fit also dey used for some special situations where communicating through talk or body movements dey difficult.

Dem fit also improve computing accessibility for people with disabilities.

For di longer term, dis research fit also help make communication between people and artificial intelligence systems more efficient.

But di real benefits of every such application must dey proven scientifically, one by one.

Dem no fit assume in advance say direct neural communication go better pass ordinary human conversation for every situation.

Human language no be just one slow tool for transferring information.

E also dey perform many important functions: thinking, reasoning, emotional expression, clarifying ideas and building human relationships.

So new technology suppose enhance people's communication capacity, no be to make dem lose dis important qualities of human language.

11. Di Scientific Contribution of Dis Research and di Path Wey Dey Front

Dis research no dey proposed on di assumption say one technology wey dey transfer all human thoughts directly don already dey created.

Instead, e dey proposed as one scientific research framework wey dey bring together di identification of brain signals, understanding di meaning of ideas with di help of artificial intelligence, measuring di performance of information transfer and evaluating people's genuine understanding.

Im important question:

No be whether di full data-transfer capacity of one electronic network fit dey created inside di human brain.

Di important question na whether di time between di moment one person form one idea inside im mind and di moment another person correctly understand dat idea fit dey reduced.

For dis one, dem must investigate whether meaning-based information representations wey dem create with di help of artificial intelligence dey provide additional performance compared with neural technologies wey dey identify ordinary speech.

E no dey enough, to say one technology dey successful, say artificial intelligence dey generate elegant sentences.

E must reduce di time wey information transfer dey take.

E must protect di true meaning of di idea wey dey sent.

E must control errors to acceptable level.

E must work reliably with different people and for practical settings.

Dis conditions must dey proven through scientific experiments.

Research must also accept say some approaches fit fail to perform better pass ordinary communication.

Di true value of scientific research no dey only for proving one expected result.

E dey for discovering, on di basis of reliable evidence, whether di hypothesis dey right or wrong.

12. Conclusion

Over di past few decades, computer data-transfer technologies don advance well-well.

Today, computers wey dey located for different parts of di world dey exchange very large amounts of information at high speed.

But communication between people still largely dey depend on talk, writing and other physical forms of expression.

Recent neural research don produce notable advances for reconstructing speech from brain signals, identifying speech wey person dey imagine for mind, and transferring defined information between brains.

All di same, one technology wey dey transfer every kind of human thought directly, accurately and at high speed never dey created yet.

E still remain one great scientific challenge wey dem go solve.

For di future, di integration of artificial intelligence and neural technologies fit create new ways to reduce di time wey people need to express dia thoughts.

Practical research for dis one must begin from di neural signals wey people deliberately dey try to convey.

Models wey dey identify dem accurately must den dey built.

Next, dem must assess how accurately di true meaning of di information dey protected.

Finally, dem must prove scientifically whether dis technology genuinely dey improve di speed and performance of communication between people.

Na only den work fit move forward towards more advanced technologies like complete direct brain-to-brain communication.

Di Ultimate Goal of My Research

Di ultimate goal of my research na to reduce as far as e dey possible di gap between di speed wey one person dey form one thought inside im mind and di speed wey another person dey accurately understand di true meaning of dat thought.

For dis one, my proposed approach na to study artificial intelligence, brain-computer interfaces, neuroscience, information theory and secure high-speed communication technologies together.

Our goal no be just to send more data at high speed.

Our goal na to build di scientific foundation for one new communication method wey fit transfer people's thoughts and ideas safely, accurately and for di least possible time, without changing dia true meaning.

References

1. Coupé, C., Oh, Y. M., Dediu, D., & Pellegrino, F. (2019). Different languages, similar encoding efficiency: Comparable information rates across the human communicative niche. Science Advances, 5(9). https://doi.org/10.1126/sciadv.aaw2594
2. Zheng, J., & Meister, M. (2025). The unbearable slowness of being: Why do we live at 10 bits/s? Neuron, 113(2), 192–204. Published online: December 2024. https://doi.org/10.1016/j.neuron.2024.11.008
3. Willett, F. R., et al. (2023). A high-performance speech neuroprosthesis. Nature, 620, 1031–1036. https://doi.org/10.1038/s41586-023-06377-x
4. Littlejohn, K. T., et al. (2025). A streaming brain-to-voice neuroprosthesis to restore naturalistic communication. Nature Neuroscience, 28, 902–912. https://doi.org/10.1038/s41593-025-01905-6
5. Wairagkar, M., et al. (2025). An instantaneous voice-synthesis neuroprosthesis. Nature, 644, 145–152. https://doi.org/10.1038/s41586-025-09127-3
6. Card, N. S., et al. (2026). Long-term independent use of an intracortical brain–computer interface for speech and cursor control. Nature Medicine, 32, 2504–2510. https://doi.org/10.1038/s41591-026-04414-6
7. Jiang, L., et al. (2019). BrainNet: A Multi-Person Brain-to-Brain Interface for Direct Collaboration Between Brains. Scientific Reports, 9, Article 6115. https://doi.org/10.1038/s41598-019-41895-7
8. Kunz, E. M., et al. (2025). Inner speech in motor cortex and implications for speech neuroprostheses. Cell, 188(17), 4658–4673.e17. https://doi.org/10.1016/j.cell.2025.06.015
9. Zhao, Z., et al. (2026). A neural architecture for imagined and overt speech motor dynamics. Nature Neuroscience. https://doi.org/10.1038/s41593-026-02456-0
10. Tates, A., Matran-Fernandez, A., Halder, S., & Daly, I. (2026). Decoding speech imagery or just noise?: A symptom of the replicability crisis. Journal of Neural Engineering, 23(5). https://doi.org/10.1088/1741-2552/aea294
11. UNESCO. (2025). Recommendation on the Ethics of Neurotechnology. https://www.unesco.org/en/node/86248

By Keshavan Thiagaraja

Powered By K7