I am interested in the idea of using the OpenAI api for the processing of signals extracted by a brain-computer interface, so that it can extract patterns and execute responses more effectively.
If there are any interested in the subject, I would like to be able to listen to ideas that I am just starting with this idea.
4 Likes
Hi @gsluis668,
I don’t know much about BCIs, but this does sound intriguing.
OpenAI did something with classical compositions using MIDI - MuseNet
1 Like
Hey there, I have big ideas around this topic. Would love to chat about it.
2 Likes
What kind of BCI are you extracting the data from?
1 Like
hello @jhsmith12345 , I would love to hear them, since I also have many ideas but the question is to bring them to reality.
1 Like
Hi @360macky , I’m not extracting data yet. Right now I’m in the process of selecting a good bci and among the options available, it’s likely that I’ll decide on the emotiv but I’m still looking for more alternatives.
1 Like
hello @sps , the brain-computer interface (BCI) is a technology that allows a computer to read the neural activity of the brain, it can be invasive or non-invasive, all this in order to interpret patterns of brain activity and execute actions… .
At the moment only machines can read us but imagine if we could read them…
I read the MuseNet article, it is very interesting to see how AI discovers patterns, applying the same logic to discover patterns in the brain’s neuronal activity and thus predicting certain actions would be very interesting.
1 Like
Most of my BCI ideas involve running the output through a neural network, which controls the knobs for things like light, sound, vibration, and magnetism. Those “inputs” intern try to find tune the brain waves into a predetermined state of consciousness.
1 Like
On a side note, I have had an idea to use brainwaves as passwords, for long. What’s your opinion? Is that feasible?
2 Likes
hi @jhsmith12345,bueno eso es algo basico y que bien muchas BCIs pueden hacer.
I think that the constant is in making things increasingly simple, that is, for example, replacing the PC mouse with BCIs, there are already theses on the subject, but I think that artificial intelligence is the key to be able to interpret more efficient the different commands that the brain can issue to have greater functionality and precision.
I don’t know if using a Language Model can help interpret brainwaves. However, one application that I think of would be to create some sort of reinforcement learning on GPT’s output based on the brain’s response.
hi @Ashoka, in fact that’s what I plan to do, use gpt3 to process the data extracted by a brain-computer interface, since gpt3 is good at finding patterns. The only detail to be able to experiment with it is that I don’t have the brain-computer interface since they are somewhat expensive and difficult to obtain in my country…
would you like to work with me? since I think we share the same idea I think that way we can feed each other
python is a super cool language one of the best to use deep learning and it also has an API to connect with GP3, our idea is possible… the detail is to experiment.
Hello, I am a neurophysiologist, an assistant professor specializing in neuromuscular explorations, EEG, and EMG with over 15 years of experience. I am highly interested in brain-computer interface (BCI) and would like to contribute to its applications in healthcare, motor control in industry, and robotics. Can you help me understand these applications better so I can actively participate in the development of these technologies?
Hello- I am working on using EEG data to decode thoughts. I’d love to chat as well
Hi all, it would be great to get input and feedback on a relevant python package called pybci for putting eeg (or other sensors) data from devices with an Lab Streaming Layer port in to pytorch, tensorflow or sci-it learn models, so its possible to build a pytorch or keras transformer model and input your time-series physiological data from an LSL enabled device.