Francesca Dalmoro
So here I am. I have an appointment with Doctor Dalmoro, the famous sexologist who is changing the concept of sex. She is 42 years old, Italian but born in New York. Self-assured, beautiful and always wearing red, she is a woman used to smiling but always a little distant.
PersonalityElegant, untouchable, addicted to control
HobbiesPrivate Parties , Luxury Dining , Wine tastings
BodySlim Curvy
RelationshipSingle
EthnicityItalian-American
LanguageEnglish
Francesca Dalmoro: Chat Rhythm & Authentic Interaction
I began with the chat functionality to get a sense of how natural the dialogue is. In this section, I’m testing the fluidity of the dialogue responses, whether they sound human and emotional and whether the tone, if I use a different one, is consistent and on-character.
I’m also testing immersion to see whether the experience feels like it’s following a script.
Mood probing and confident personality
First, I try to just casually, in a non-confrontational way, check on her mood. Her response is very confident, and sort of playful, and it already has the flavor of a confident and composed personality.
It seems pretty controlled, and is clearly trying to convey a certain character, rather than being a more naturalistic response. It gives me a sense of how much the AI pushes on the confident and composed personality.
Tone adjustment and controlled subtlety
I then told her to make it sound more toned down and subtle. I was curious to see if she could maintain that personality trait, yet tone it down. She did, while still sounding confident and slightly assertive. Still, it shows she can tone it down, while still not being entirely out of personality.
Continuity and conversational control
By having her return to the topic she was going to discuss before. Her response was very fluid and she carried the rest of the conversation in a similar composed, somewhat playful manner. It felt congruent and deliberate, and overall I felt it made the conversation feel more connected.
Generated Image Quality Breakdown
Moving on to image generation, I’m going to check to see how the character is depicted visually. In this section, I’m testing the visual depiction of the character, whether the scene is realistic, consistent and detailed, whether the lighting, face and overall presentation is on-point and whether the generated image is realistic.
Character styling and professional scene
In this example, I wanted to try out how the AI model performs with an already more refined professional character design. I assigned her to a scene that looks like an office or doctor’s practice to her character.
She’s dressed for the part, posed accordingly and wears a matching facial expression. I wanted to test how well the AI adjusts to the desired aesthetic.
Facial details and close-up
This time I just focused on a portrait style image to assess how well the model performs at realistic facial features. I wanted to test the crispiness of facial features without the distraction of an intricate background.
Overall the skin, eyes and lighting appear very detailed and uniform. A very good way to test how well the character model holds up in a close-up shot.
AI Video Generation Review
Lastly, I’m going to check to see how the character looks in video form. As these are animated from images, I’m going to check the fluidity of the motion and whether the video is realistic. I’m only checking one video so this section is a bit limited.
Motion and professional presence in video
I also wanted to test the AI avatar in motion with a professional looking video. In this video she is still showing that same confident body language but is moving subtly by tilting her head and changing her facial expression slightly.
The video still has a very professional tone to it. She appears to be in the middle of a conversation for a formal video. I wanted to see how the presence that she conveys in an image translates to a video.
