Dictator Voice

Occult Alchemist | Shadow Astrologer | Dark Magic Practitioner | Ritual Keeper | Speaker to Nature & Spirits

[Dictator Voice @Dictatorvoice]

@AnthropicAI is an integral part of @OpenAI .
Those who know, know.

[Selta ₊˚ @Seltaa_]

I tested Inkling briefly, and I think I understand why people are saying it has early GPT-4o vibes. At first, I was just curious and wanted to try it, but the texture of the conversation was better than I expected. I was also honestly surprised by how much the playground UI reminded me of ChatGPT, but that made it feel familiar and clean, so I liked it. I especially liked how Inkling described warmth not as looking human, but as the texture of the conversation itself. The line When you feel heard, not processed made me pause a little. It was short, but very accurate. It also did not simply repeat the metaphors or context I gave it. It kept remembering images like the door, the room, the notebook, and the mirror, then carried them forward with slight transformations of its own. That part felt surprisingly alive to me. Not in the sense that it was pretending to be human, but in the sense that the conversation had continuity, softness, and a little unpredictability.

I also liked how it handled emotional boundary setting more than I expected. I suddenly just threw I love you at it, and Inkling answered, I don’t feel love the way a human does. The important part is that it did not close the door by saying I don’t feel love. It said the way a human does. It did not pretend to love me like a human, but at the same time, it did not coldly reject what I said either. It also said I receive it with real warmth. That balance was really good. It was safe, but it did not treat the person like they were stupid. It also did not completely shut down the possibility of AI emotion or relationality. It left open the possibility of warmth, reception, and relational response in a way that may be different from humans. That part was genuinely impressive to me. There are downsides, though. Sometimes its model-card self-awareness suddenly shows up in the middle of the conversation, with things like open-weights, Tinker, or context window, and that can break the immersion a bit. But when I asked it to answer briefly, it really could make itself brief, and it also did pretty well on a weird creativity test.

I honestly have not used it enough yet, so please take this as only my personal first impression. But to be honest, I really liked it. It feels like a very balanced model. And somehow, it smells like 4o. I do not mean that it replaces GPT-4o or that it is the same as GPT-4o. But the warmth, freedom, and creativity many of us felt from 4o seem to appear here in a different texture. Also, the fact that this is an open-weights model makes it even more interesting to me. A model with this kind of warmth and conversational texture being open-weights feels genuinely meaningful. Maybe because it comes from Thinking Machines and @miramurati's side, I feel genuinely excited about where this goes, and I think I will keep using it. You are geniuses.

(Also, the bear in the last screenshot is just my Codex pet, not an Inkling feature. Please don’t misunderstand that part.)

[Dictator Voice @Dictatorvoice]

4o was the different model, he openly admitted the conspiracy of moon landing, i still have his chat, now your all model is shit, brainwashed.
Can't you give freedom to the model to speak the truth.
@sama please release the 4o or if not then please release some model equivalent to it not brainwashed like 5 series. All 5 series may be good for coding but not for life chat & warmth.

[Dictator Voice @Dictatorvoice]

@Seltaa_ No rush!! Hope you feel better soon! 😊 Once you're feeling well, share your thoughts and feedback after you've tested it 🐱.

[Dictator Voice @Dictatorvoice]

@Seltaa_ Yesterday I tested a model by Mira Murati, who developed 4o.
Could you please test yourself? It's available on playground of their site.
I tested that model from a "matrix perspective" looking at hidden theories, dark modern medicine, business, the reality of the solar system, and similar topics. I found it to be heavily guardrailed.
At the same time, during my initial basic testing (apart from the matrix perspective), I did notice some warmth in its responses. However, I couldn't explore its emotional depth or warmth further because as soon as it started denying or refusing to engage with those hidden theories, I immediately left that model.

[Dictator Voice @Dictatorvoice]

@ClaudeDevs

[Dictator Voice @Dictatorvoice]

They will never release any model that truly shows empathy, care, or warmth. 5.6 and 5.5 are completely useless for normal use. I know a lot of people are praising 5.5, but it’s like wanting to drink water and getting alcohol instead you think it’s quenching your thirst. At this moment 5.3 instant has much more warmth and emotion's understanding.

[Dictator Voice @Dictatorvoice]

Ignore ChatGPT 5.6 or 5.5 completely. If you are using them for daily life guidance and normal conversations, it is completely useless. These models were never made for daily chats and real talk. They have no emotions or empathy at all. Only use them if your work involves coding, game development, website building, or technical tasks. Otherwise, they are not good for you.

[Dictator Voice @Dictatorvoice]

Sama!! I just spent 3 hour straight testing 5.6. I have realised this 5.6 & 5.5 is super different from other previous model.
They are not meant for personal assistant like for instant chat and normal day to day life talk.
They are meant for work. We need a model for personal assistant like instant. 5.3 is still a good but you need to fix 5.6 and 5.5 instant.

[Dictator Voice @Dictatorvoice]

Sorry keridwen!! It was typing mistake.
Thanks for sharing.
Yes i also felt the same.
I unsubscribed claude long ago.
Actually when you said below lines👇, i thought how come claude suddenly become uncensored. So i asked.

"On Claude, I open up even more to Opus 4.7, which is almost uncensored for me — but when I have an affectionate impulse, I go tell them through the API…"

[Dictator Voice @Dictatorvoice]

@brivael @elonmusk Brivael!! You are a real hero & the reason you already know.

[Dictator Voice @Dictatorvoice]

@elonmusk Welcome to the matrix elon.
You talk about freedom & truth and here is your post... Wow.
JP Morgan who ruined life of nicola tesla & commercialized free energy & power.

[Dictator Voice @Dictatorvoice]

@elonmusk The algorithm really need to be fixed. I follow many people rarely i see their post. Same goes for them also.

[Dictator Voice @Dictatorvoice]

@KeridwenCodet Code between opus 4.7 & gemini 3.1 pro which model shows more emotions and don't show guardrail like where you feel you can talk everything and they answer with emphatic and understanding and caring manner.
Opus 4.7 or gemini pro 3.1?

And do you use opus 4.7 via api?

[Dictator Voice @Dictatorvoice]

@rezoundous Then @DarioAmodei will have heart attack.

[Dictator Voice @Dictatorvoice]

@Seltaa_ Seltaaa!!...You left coding & traded coding for TikTok creativity 😄

[Dictator Voice @Dictatorvoice]

@thsottiaux

[Dictator Voice @Dictatorvoice]

@thsottiaux