[ot][spam][crazy] eeg thoughts

[k]

preface: hobby eeg research is very old.  but it's not something you
see people talking about a lot, despite there being public information
on making devices and software, and cognitive amplification.  it needs
a little extra protection.


say whenever you tried to do something, your body did something.
if you wanted to form a computer algorithm to detect this, the current
approach is called "machine learning".  it works best with many
thousands of data points.  this means marking in data, many thousands
of events when your body does something, when you try to do the thing.
you could then train an algorithm to predict it.  i don't know the
actual number needed.

there may be approaches that take only a few datapoints, but they
would still be split into a train and test set.
additionally there are approaches that are less automated.

another interesting avenue is using pretrained models that have
emerged from research.  these models already report brain or event
metrics, and could possibly be combined to chart the behavior of a
brain.  there are communities where researchers chat with hackers
about these things.  one such community is neurotechx, which is on
slack.  here's an invite if somebody else is reading:
https://join.slack.com/t/neurotechx/shared_invite/zt-dfmxe9r8-2TnyN598LS7ZrvJB4kZroA

it's quite clear there are older hacker communities.  i'm not sure
where to find them, but one could look on irc servers for interesting
room names or usernames of developers of say the openeeg project or
the datalad project.

--
pasting the invite was a little hard, sending now :)