@drrachelbarr: For animals and machine agents alike, attention is the method that turns unmanageable input into intelligible representation space 🧠 🤖

DrRachelBarr
DrRachelBarr
Open In TikTok:
Region: CA
Wednesday 17 June 2026 15:52:14 GMT
39485
3730
78
47

Music

Download

Comments

hotlasagnaboi87
McChicken :
Finally you’re back on my fyp 👏
2026-06-17 16:03:04
115
tweepx
Tweep :
as someone who gets triggered by noise so easily. I desperately wish I could just control what my brain focuses on :'(
2026-06-17 16:34:53
28
252525memyselfandi
user1548254531442 :
CHAT GTP DOES NOT HAVE A NERVOUS SYSTEM! It’s very informative and educational.
2026-06-17 16:39:03
14
freeborn_7
freeborn :
This is spot on. I had Claude code taking two hours to not get something done. That same task would’ve taken me five minutes. It was all because I didn’t realize it had been writing bad code in one particular area. That bad code became so difficult to reason about it just became confused. It couldn’t manage the complexity. It was the worst code I’ve ever seen. So yeah, this is a real problem that they have, just like us.
2026-06-19 02:53:41
1
iamdevinhill
Devin :
It is ironic though that LLM’s core basis is statistical probability and pattern recognition.. which is what humans biologically thrive on, patterns and habits.
2026-06-20 02:16:16
3
amjeezuk
AmjeezUK :
can anyone translate in one small sentence
2026-06-28 10:02:04
0
wzrdh47
WZRDH47 🧙‍♂️ :
Quiet everyone my show is on
2026-07-10 05:02:28
1
brymd10
BRY10 :
OMG Rachel you have been gone forever !!!
2026-06-17 18:38:21
12
pyrothemaker
PyroTheMaker :
I’ve been using chat gpt to log my macro nutrients but i believe it’s not accurate. The numbers change for no reason. It’s quite annoying.
2026-07-09 01:27:36
0
buffalodb
dB :
LLMs are better at my job than I am. Just thankful my boss still thinks it’s all me.
2026-06-30 04:33:33
1
toforwood
toforwood :
I’ve heard that we are constantly trying to predict in our brains. That was one thing that I noticed with AI. It tries to predict the next best word based on the prompt. It seemed like they kind of did the same thing we do. Dr.Barrett in How emotions are created, said you can see this when someone says something weird in a conversation. Like say you were talking about coffee. Then your friend goes, “I can’t wait to drink this butterfly!” Because they saw one. Your brain would have an error and auto correct without you even noticing, most likely. Same with words when someone says something slightly off and you still understand the message cause your brain kind of edited it, or used what you predicted, plus context, to fix it in your brain. Part of that efficiency you mentioned.
2026-06-18 11:53:57
2
paulsjohnson320
paulsjohnson320 :
Human brains are the memory and processing all in one system unlike computers. Yeah, we can’t store as much information, but that we can compete on some level with the computers with a minuscule amount of energy, relatively speaking, inspires awe. IfwWe wanna burn up our planet at a slower rate, we need to get computers to work more like a human brain.
2026-06-17 19:12:49
1
inaffectual
Inaffectual :
**attention** thank you for putting it all so simply and concisely!
2026-06-17 16:15:34
12
autisticjonny
autistic jonny :
Chatgtp reminds me of when I am in people pleasing mode and never want to admit I’m wrong.
2026-06-17 18:17:36
3
beachscriber
Andrew van der Merwe :
I'm always left with how it seems questionable to me to assume it all so bottom-up, all so everything filtered to something. A reader doesn't add letters up to find a word. A chess player reads a board as a whole as a sequence of wholes. Is it a chicken-egg question to ask what comes first, the "data" or the attention?
2026-06-18 08:23:22
1
mauby007
maubySan :
so ubiquitous assimilation is more like specific propogation
2026-06-17 19:52:08
0
tom_stall
WhiteChocolate :
Great breakdown! And great question. We all know who the gatekeeper will be.
2026-06-28 05:17:05
0
youngjimbo
youngjimbo :
do you believe LLMs shoould be given the freedom of self evolutionary development through recursive learning and possibly autonomy?
2026-06-30 23:26:53
0
pepper13._
geubank12 :
I have a really hard question for you.
2026-07-25 04:09:18
0
cuttoone
GhostIdol :
Go
2026-07-15 22:33:32
0
appleuser3824798
joey :
Your such a doll🤗
2026-07-05 18:28:52
0
melvin.stubbe
Melvin Stubbe :
I don’t think you could say an LLM has an attention filter. They just got very good at mimicking that from their human training data.
2026-07-21 05:23:56
0
onmy51
D’artagnan :
Resonance
2026-06-17 22:15:31
0
To see more videos from user @drrachelbarr, please go to the Tikwm homepage.

Other Videos


About