Two questions about the state of affairs

[Onwrikbaar]

Please satisfy my curiosity

1. What, in your perception, have been the most meaningful innovations related to electrostimulation in the last decade or so?
2. What innovations related to electrostimulation would you love to see before the year 2030?

Thanks!
Mark

[NotARealName]

I was actually going to post today about this, I'll still do it.
in short AI. See my post in a little bit.

[admin]

I really hate this obsession with AI, its like a few years ago when everything was BlockChain/BitCoin.

The processing required for AI is massive and at the moment its a bit of a buzz word... people are almost looking for something to apply AI to, I'm just waiting for the AI controlled toilet door opener.

Si

[Riddle]

I am not an AI fan. Why use so much data and processing power to get some vague results? Why not just give the computer clear, concise, and precise instructions to get the exact results desired?

I think the computer controls are the most significant part and that is where I want to see the most significant advances. Would love to see more precise automation and control of the stim experience. At the same time, I want a better “random” mode for adding pain and pleasure into self-bondage sessions.

[estim4fun]

The AI gives answers, that does not always fit your question.
An example says the AI does not work. They places an ox on a supermarket parking, the system does NOT detect it as an ox. The reason is the AI correlates with a field. It is not wrong, a farmer who uses a large number of CCTV may want to have an automatic counting of his livestock in his fields to ensure no one is missing. In such case the system works (and throw an alarm)

[Onwrikbaar]

A couple years ago I attended a little seminar by Edge Impulse, a company specialising in Machine Learning for resource-constrained devices. I was quite impressed, and the following week I played a bit with one of their demos - wakeword recognition running on a cheap Nordic development kit (using a nice STM sensor board with a microphone among other things). You train the model on your desktop computer and upload it to the development kit's embedded controller, where it is then executed. It ran very well on a 64 MHz controller with single precision floating point unit - which is considered low-end by today's standards.
AI may be an annoying hype, but it can do amazing things and we are definitely going to see a whole lot more of it, even on cheap and battery-powered hardware.

For a while I thought it would be nice if an e-stim box could recognise a few simple words (if only "STOP"), so you wouldn't have to reach over to adjust it with slippery hands But if a power box can connect to a computer or phone, you may as well use those for functions like speech recognition and to control the box through its API.

[estim_si]

I am not an AI fan. Why use so much data and processing power to get some vague results? Why not just give the computer clear, concise, and precise instructions to get the exact results desired?

I think the computer controls are the most significant part and that is where I want to see the most significant advances. Would love to see more precise automation and control of the stim experience. At the same time, I want a better “random” mode for adding pain and pleasure into self-bondage sessions.

Better random…. Longer? harder? Can you give me some parameters that would work for you?

Si


Sent from my iPhone using Tapatalk

[Riddle]

I am not an AI fan. Why use so much data and processing power to get some vague results? Why not just give the computer clear, concise, and precise instructions to get the exact results desired?

I think the computer controls are the most significant part and that is where I want to see the most significant advances. Would love to see more precise automation and control of the stim experience. At the same time, I want a better “random” mode for adding pain and pleasure into self-bondage sessions.

Better random…. Longer? harder? Can you give me some parameters that would work for you?

Si


Sent from my iPhone using Tapatalk

I have been thinking of how to describe the random mode I want since yesterday and think that I have figured out how to describe it. Instead of the random mode on the 2B, I would like a random mode in the Commander program that picks 2B modes and settings at random and then runs it for a short period also picked at random within the user setup instructions. So, instead of random pulses like the 2B random mode, the 2B settings are picked randomly and the user will not know what settings are selected next.

[admin]

that could be done in something like python - the only issue would be deciding what settings you want in each mode... randomly choosing pulse mode at 10% and waterfall at 50% are two very different things

Si

[StimJoy]

....I'm just waiting for the AI controlled toilet door opener.

https://www.businessinsider.com/kohler- ... exa-2024-1

Ok, maybe not quite AI in the current hype sense...but it does open the lid for you and you can apparently talk to it.