Brain-Com­puter Interfaces

TagLast edit: 24 Nov 2021 5:14 UTC by Multicore

A Brain Computer Interface (BCI) is the generic term used to describe any kind of system that serves as a communication bridge between the brain (human or not) and an artificial module. It’s a field of research in which wide investment has been made since the 1970’s, especially in the clinical fields and ergonomics. Generally speaking, any kind of brain activity that can be recorded can be used as a means of communicating with another system. Through the use of statistical classification techniques it’s possible to associate certain states or characteristics of the recorded signal – which the experiment subject learns to control—to any procedure, usually mediated by a computer.

Many techniques have been developed to help us look and better understand the way the brain works. They range from imaging techniques (like MRI, fMRI, fNIRS or PET), to electrophysiological ones (like EEG, EcG or MEG). While the first category is usually used to obtain high resolution images of brain structures and the second one to register and analyze the electrical activity produced by the brain, with a high temporal resolution – which is why they are the ones mainly used in the field of BCI’s. In pair with such methods, although a different area in itself, includes brain implants capable of communicating directly with the neuronal tissue—neuroprosthetics.


Of all the different means avaliable, the registering of the electroencephalographic (EEG) activity is the most developed and extensively researched of this fields. It allows us, in a non-invasive way, to peak the brain functioning with a high temporal resolution – furthermore, it is now well established that different brain states produce distinct observable activity. With the help of electrodes placed on the scalp, it is possible to feed this activity and their respective variations and patterns to any system capable of classifying and detecting them in real time and act accordingly (making this a field highly interconnected to that of machine learning).

The field of BCIs has followed closely the developments in signal processing and classification, along with the increasing computational power available. It was firstly researched as a communication means (for people unable to move, for instance) through the detection of ERPs – event related potentials, small variations of amplitude associated to the presentation of certain stimuli—as well as a way of automatically detecting epileptic seizures. Also, much owing to the first and major financers of such research, the DARPA, the use of BCIs has been always closed associated to the military field. This has allowed insights regarding the detection of mental states of fatigue and attention variations, which has led to the development of informatics systems capable of adapting to the mental state of the user.

Currently we have available a considerable range of both research and commercial applications of EEG based BCI systems with a wide list of applications. It has shown to be a field due to receive increased attention in the next years, especially through the developing of increasingly efficient classification algorithms and computer power, and the fascination with the cognitive augmentation it might bring.

Potential applications

Although the EEG has been the main technique used for the development of such systems, it has been shown to be possible to integrate electronic controllers directly in the functioning of single cells or even networks. The permanent implant of devices for interpretation and regulation of cortical activity has also been demonstrated.

This has led to a renewed interest in the field and the exploration of new hypothesis, like drug rehabilitation through the detection of relevant cues and stimulation of the brain reward system, rehabilitation after strokes or lesion and even direct transmission of patterns of thought between subjects.

Other attractive future application includes the upload of the whole content of the brain, and thus the mind, to a computer. Although still speculative, it seems theoretically possible.

External Links

Further Reading & References

See Also

Brain-Com­puter In­ter­faces and AI Alignment

niplav28 Aug 2021 19:48 UTC
35 points
6 comments11 min readLW link

The ‘Ne­glected Ap­proaches’ Ap­proach: AE Stu­dio’s Align­ment Agenda

18 Dec 2023 20:35 UTC
163 points
20 comments12 min readLW link

Us­ing Brain-Com­puter In­ter­faces to get more data for AI alignment

Robbo7 Nov 2021 0:00 UTC
43 points
10 comments7 min readLW link

We have to Upgrade

Jed McCaleb23 Mar 2023 17:53 UTC
126 points
35 comments2 min readLW link

Assess­ment of AI safety agen­das: think about the down­side risk

Roman Leventov19 Dec 2023 9:00 UTC
13 points
1 comment1 min readLW link

cy­ber­punk raccoons

bhauth28 Apr 2023 2:52 UTC
20 points
7 comments5 min readLW link

Idea: NV⁻ Cen­ters for Brain Interpretability

James Camacho18 Feb 2024 5:28 UTC
10 points
1 comment3 min readLW link

When does tech­nolog­i­cal en­hance­ment feel nat­u­ral and ac­cept­able?

Gunnar_Zarncke1 May 2015 21:11 UTC
8 points
34 comments5 min readLW link

BCIs and the ecosys­tem of mod­u­lar minds

beren21 Jul 2023 15:58 UTC
88 points
14 comments11 min readLW link

[Question] What are we pre­dict­ing for Neu­ral­ink event?

Dr_Manhattan12 Jul 2019 19:33 UTC
32 points
15 comments1 min readLW link

Watch Elon Musk’s Neu­ral­ink presentation

Alexei19 Jul 2019 21:48 UTC
9 points
4 comments1 min readLW link

Brain-Brain communication

Jordan9 Dec 2011 17:05 UTC
12 points
22 comments1 min readLW link

[Question] Thoughts on Neu­ral­ink up­date?

NoUsernameSelected29 Aug 2020 22:59 UTC
14 points
5 comments1 min readLW link

[Link] Tools for thought (Ma­tuschak & Niel­son)

ioannes4 Oct 2019 0:42 UTC
20 points
7 comments2 min readLW link

Gra­di­ent Des­cent on the Hu­man Brain

1 Apr 2024 22:39 UTC
46 points
4 comments2 min readLW link

Re­search Direc­tion: Be the AGI you want to see in the world

5 Feb 2023 7:15 UTC
43 points
0 comments7 min readLW link

[Linkpost] MindEye2: Shared-Sub­ject Models En­able fMRI-To-Image With 1 Hour of Data

Bogdan Ionut Cirstea10 Mar 2024 1:30 UTC
10 points
0 comments1 min readLW link

[Cross­post] In­tro­duc­ing the Hyper­man­i­fest: Redefin­ing AI’s Role in Hu­man Con­nec­tion and Interaction

Suzie. EXE7 Apr 2024 17:21 UTC
4 points
0 comments5 min readLW link

Fore­sight In­sti­tute: 2023 Progress & 2024 Plans for fund­ing benefi­cial tech­nol­ogy development

Allison Duettmann22 Nov 2023 22:09 UTC
24 points
1 comment6 min readLW link

Distil­la­tion of Neu­rotech and Align­ment Work­shop Jan­uary 2023

22 May 2023 7:17 UTC
51 points
9 comments14 min readLW link

Scien­tists make mon­keys smarter us­ing brain im­plants [link]

Dreaded_Anomaly15 Sep 2012 18:48 UTC
32 points
4 comments1 min readLW link

The Neu­ral­ink Mon­key Demo

scroogemcduck13 May 2021 2:29 UTC
19 points
9 comments1 min readLW link

SRG 4: Biolog­i­cal Cog­ni­tion, BCIs, Organizations

KatjaGrace7 Oct 2014 1:00 UTC
14 points
139 comments5 min readLW link

[Question] Creat­ing su­per­in­tel­li­gence with­out AGI

Antb17 Oct 2022 19:01 UTC
7 points
3 comments1 min readLW link

Without—MicroFic­tion 250 words

Carissa Cassiel9 Dec 2023 21:49 UTC
19 points
1 comment1 min readLW link
No comments.