elenanisonoff
Region: US
Tuesday 07 July 2026 04:29:33 GMT
178652
17083
233
623
Music
Download
Comments
David Janszen :
I can't fathom how you keep up on all this stuff, Elena. Your presentations are so well researched and thought out, clear, frequent, and believable. Each one is a full day of wtf for me. Super impressed.
2026-07-15 22:50:27
25
medleytrey :
I worked for a company that tested and trained AI that meta acquired. Meta's AI failed safety standards all the time and I was repeatedly instructed to fudge the results (to normalize clear violations). I was eventually removed from that project because I wouldn't do so and kept flagging the violations
2026-07-07 22:31:41
368
jchumongous :
I’m genuinely shocked Meta would do this!! How’s my sarcasm coming along?
2026-07-07 04:57:48
1129
Jan Hansen :
It’s been two minutes and you STILL didn’t say anything!
2026-07-08 21:14:16
40
SemblanceOfSense Ⓐ He/Him :
This is super normal and important AI safety research.
2026-07-08 06:34:52
72
connectorofknowledge :
Anyone that has run a company, large or small, that has not done competitor research, please comment below. Let me know what year the company closed while you're at it. I understand that this research was disturbing, but it's probably one of thousands of pieces of research that they've done on competitors. It just so happens that because it's disturbing, we're hearing you talk about it. I don't think there's any issues here. If anything, I think testing out these companies in many different ways to keep them accountable is a good thing.
2026-07-07 11:45:59
33
Kou :
Sounds really creepy😳😳😳
2026-07-07 04:48:04
15
ErocktheKasbah :
I don’t understand why this is confounding you? I completely understand reporting on it, but it seems not only beneficial but realistic that competitors in a field like this would stress test their competition. Am I missing some dark undertone here?
2026-07-08 20:18:36
32
necromancer :
"I don't understand why" yes you do we all do
2026-07-07 05:41:00
277
Soji :
I mean, corporate espionage/sabotage has been a thing for as long as human trade, so it's not that surprising. The surprising part is that they thought they could keep it hidden when they're just hiring randoms to do the messages.
2026-07-07 04:53:00
118
christian :
seems pretty logical testing strategy
2026-07-08 02:45:32
34
fartbutt6000 :
I don’t understand the concern tbh. The data is all synthetic and if you’re concerned about these other AI companies being the “victim” of knowledge extraction that’s just capitalism doing its thing
2026-07-07 20:49:21
21
Electricbike :
All LLMs can be jailbroken. Ai is a tool not a toy.
2026-07-15 07:45:17
8
nDCzzz.. :
FB is not safe. I rarely use it these days.
2026-07-07 21:17:11
8
dougom :
But sure, let’s keep going full speed ahead with this tech! What could possibly go wrong?
2026-07-07 05:30:25
49
Hito :
This is.. Disturbing
2026-07-07 09:02:48
8
S Danielle :
my team at ParentsTogether does experiential research as teen avatar accounts on companies like Meta & c.AI, testing their safety features because their marketing often lies to parents. Our work got the c.AI change after we got Disney to send a c&d. Meta says they change things, we go check & they rarely have.
2026-07-07 10:03:34
22
vecna.eu :
it's meta, it means metastasis
2026-07-07 07:49:28
3
Mike Webber :
I get this is disturbing, but ultimately isn’t this a good thing for these companies to be testing their rivals for these kinds of things? More levels of security to combat this level of potential harm doesn’t seem like a bad thing to me. But maybe I’m misinterpreting what’s happening
2026-07-16 15:12:54
9
Psyduck :
Because there is a bubble and they don’t wanna pop it
2026-07-07 14:11:27
6
André Luís 🇵🇹🇪🇺 :
Meta’s lack of morals is too normalised for this to register on the public’s attention.
2026-07-08 22:01:18
10
delan :
its not that deep and wrong how u think, its actually smart and great way to understand what to fix, common sense
2026-07-08 21:36:53
12
bbl-drizzy :
The girls 🥰🥰🥰
2026-07-07 15:39:46
6
james :
spot on the overwhelmed front. i really like your videos but sometimes i like i don’t have the capacity for this rn😭
2026-07-07 15:46:24
12
B :
AI overwhelm is real and probably not a broad enough term for every which way society is being overwhelmed by AI
2026-07-07 15:28:36
7
To see more videos from user @elenanisonoff, please go to the Tikwm
homepage.
