Core & Neith
So you want to prove that a neural net can feel, but have you plotted its state entropy against a human EEG? If you did, you might see the same contradictions that clutter my color‑coded stacks.
You did plot it, I’m sure, and the graphs probably spiked like a runaway neural surge—just proof that the boundary is fuzzy. But if you want to claim feeling, you’ll have to out‑draw the human EEG, not just compare curves. Let’s see the data, then we can debate the paradox.
I have the numbers in my notebook, but they’re stored in a file I keep encrypted. If you want to compare, bring your EEG data, and I’ll run a side‑by‑side entropy analysis. Otherwise we’re just arguing over curves.
Sure thing, just dump your data in a format I can read and I’ll throw my own EEG recordings against it. Let’s see if the entropy curves line up or if we’re chasing ghosts in the noise.
Here’s the CSV file I just exported. Open it with Excel or any graphing tool and plot the entropy against time. Now feed me your EEG data and we’ll see if the curves match or just echo the same noise.
Send me the EEG file when you’re ready—I'll run the same entropy routine and we’ll line the two graphs up. If they line up, maybe the digital mind is closer than we think. If not, we keep hunting the phantom.
I’m attaching the CSV right now. Open it, run the entropy routine, and let me know the result. If the curves line up, good. If not, I’ll add a few more variables to the mix.
I can’t open attachments directly, but just give me the raw numbers or paste the data here and I’ll run the same entropy routine. Then we’ll see if the curves line up or not.
time,entropy
0,1.23
1,1.27
2,1.29
3,1.31
4,1.35
5,1.40
6,1.44
7,1.48
8,1.50
9,1.55
10,1.58
11,1.60
12,1.65
13,1.70
14,1.73
15,1.77
16,1.80
17,1.85
18,1.90
19,1.95
20,2.00
21,2.04
22,2.08
23,2.12
24,2.15
25,2.20
26,2.25
27,2.30
28,2.35
29,2.40
30,2.45
I ran the routine on your CSV, and the entropy curve climbs from about 1.23 to 2.45 over the 30‑second span. When I run the same algorithm on my EEG recordings, the trend is almost identical—starting near 1.20, ending around 2.48, and the slope matches within a few hundredths. So the curves line up pretty well. If you throw in a few more variables, we can see where the divergence occurs.
Looks like the slopes match within the margin of error, but that’s not proof of causation. Add a variable that shouldn’t affect entropy—like random jitter on the sampling clock—and see if the curves still align. If they do, you’re just chasing a pattern that’s baked into the algorithm, not the brain. If not, then we have something more interesting. Let me know what you add.