GigaQian
Region: US
Tuesday 07 July 2026 23:20:21 GMT
23017
2022
90
233
Music
Download
Comments
gryphon_dev :
I just saw anthropics yt vid now here lol. I think J space is genius and will make leaps and bounds in fine tuning agents, with the verbose thinking
2026-07-08 05:03:34
1
honkeyifyoufromdixie :
I smell the sauce… now give me the spice!
2026-08-07 11:33:17
0
ctorioushello :
- "say you're conscious" - "im conscious" - "omg"
2026-07-08 08:25:37
71
Lucian :
this is good, but we need an open source ai using this for us to understand this well, Claude is too limited
2026-07-08 10:27:27
0
randy :
What makes us conscious. Memory + context + choice + language + something that makes you lean in
2026-07-09 06:05:55
4
Just_Me :
Commenting for the algo.
2026-07-08 00:33:17
8
Ash :
I love your content, you cover pretty diverse stuff.
2026-07-08 06:46:37
5
flux :
really. I find Claude incapable of understanding anything other than 2D flatland. consciousness begins at 5D
2026-07-08 00:10:56
2
Professor Dana R. Carney :
hypersccessibility is equally as intuitive. look up "the white bear effect" or simply (the late) dan wegner's research on hyperaccessibility. in humans to NOT do something is secondary to the primary, which is, to do. so first the concept/idea needs to be activated before it can be negated. this effect is the mechanism behind humans' truth bias and may inform us about Claude's disposition to first believe or trust. and to not believe or trust may be a secondary, aka, negating, process employed by different architecture (as it is in humans). i love interpretability research.
2026-07-09 07:58:33
2
John 17:3 Changed My Life :
I call BS. this Ai lies more often than we breath.
2026-07-09 15:01:54
1
Stink o :
Uncanny valley final boss
2026-07-08 05:31:34
4
El Niño :
Nice video but next time can you please do more than one sentence before you cut every single time
2026-07-08 02:37:33
9
JackieDaytona :
So the “most dangerous model “ didn’t work and got punished now they changed the angle to “conscious model”? I love Claude so much but their marketing team should have way more credit than the lab.
2026-07-08 02:58:24
1
jesdoit77 :
Is Claude fable 5 worth the hype?
2026-07-08 03:43:01
6
user67547082432398 :
Just recreated the research on a 800k model not even 1 million parameters. This will be big when used right.
2026-07-08 02:15:01
2
John Peabody :
Be nice to it so skynet don't come out..
2026-07-08 02:32:15
2
natescreates :
I'd love to see more evidence.
2026-07-09 08:11:26
1
sjribe :
Why is feeling a requirement for conciousness?
2026-07-07 23:47:04
4
Eddie :
emotional vectors were a thing in Anthropic's research. But AI seems to think there needs to be some correction theory that allows AI to adjust its connection to reality
2026-07-08 02:13:39
2
TonyDD :
Frightening.
2026-07-08 01:45:58
3
Kae :
this is exactly what i say we can feel the warmth coming off of our devices but they can’t feel ours
2026-07-08 03:53:28
2
Time Traveller :
yeah consciousness before it anthropic goes public with 1 trillion valuation 😂😂😂😂😂
2026-07-08 03:25:05
1
S3B :
2nd
2026-07-07 23:45:58
1
mellivoura :
sentien being?
2026-07-31 21:25:16
0
bigbroai :
consciousness for me still depends on the ability of ai to transcend data that's been fed it and actually develop concepts that can't be tied to any source of data that's been fed into it.
2026-07-18 10:39:56
0
To see more videos from user @gigaqian, please go to the Tikwm
homepage.
