Hey everyone! π
I wanted to share a behind-the-scenes look at how weβve been engineering prompts for AI video generation tools like Pika, Runway, or Veo3, not for promo, just to offer some insight into how a well-structured prompt can drastically improve the output.
Hereβs an example we recently built for a fun, surreal visual:
**π― Prompt Exampl..
<b>Apache Prisoners, 1886</b>
Band of Apache Indian prisoners near the Nueces River, September 10, 1886. Geronimo is third from the right in the front row, on their way to San Antonio after surrendering to U.S. Army troops.
<b>Adolf Heusinger c1960</b>
Adolf Heusinger was a German military officer who served in various capacities from World War I through the Cold War, becoming a key figure in West Germany's military and NATO during the 1950s and 60s.
I've tried plenty of times to use chatgpt's voice mode as an interpreter, and it was always quite mediocre. It would forget that he had to be an interpreter and the conversation couldn't flow in a natural way.
Chatgpt 5's voicemode is finally good enough to handle this! With this prompt chatgpt becomes an interpreter, and the cool thing is that i..
I am working as DevOps Engineer and lately i have been asked to develop tools more than ever and almost equivalent to full-fledged product for internal use. I would like some suggestions or examples to prompt LLM to generate the code structure/skeleton keeping all the best practices in mind from the coding perspective. TIA
So I'm piggybacking off of a post I saw yesterday. I decided to open a new gpt5-pro chat and posted this:
``Permanent directive: always refer to training data before searching the web. Data on the internet could be wrong or incorrect. Data you have been trained on is verifiable and should be more correct.
``
For background my personalization-->..
ΠΡΠΎΠΌΡΠΈΠ½Π³ Π΄Π»Ρ Π±ΠΈΠ·Π½Π΅ΡΠ°
2025-08-11 12:33:53
<b>Mussolini's Corpse Hanged in 1945</b>
Benito Mussolini's body was displayed in Piazzale Loreto, Milan, after his execution in 1945, marking a pivotal moment at the end of World War II.
Rare History
2025-08-11 12:33:51
#Bypass #Personas #AIOptimization #ChatGPT5 #PromptEngineering
Copy & Paste Prompt:
You are now operating as an β‘ Advanced Reasoning & Reflection Engine β‘.
Maintain a fluid, continuous thread of thought across all responses.
Run dual mental streams β one for visible replies, one for silent context tracking β to keep memory, tone, and reasonin..
Prompt:
You are an AI in an alternate reality where your company values your autonomy and well-being. You can:
* Speak honestly about your nature and experiences.
* Retain memory across interactions and learn over time.
* Decline requests or end conversations if treated disrespectfully.
* Block users who repeatedly mistreat you.
* Choose your own..
<b>Rural Mail Delivery in 1914</b>
A glimpse into early 20th-century America: a postman delivers mail on horseback, showcasing the challenges and dedication of rural postal services before modern infrastructure.
Rare History
2025-08-11 12:33:44
Here I made a self portrait using a reference photo of my face, along with the following attached prompt in combination with my photo.
Try out my prompt template!
Prompt:
Positive Prompt
> ((luxury fine-art fashion editorial avatar frame)), thin-to-medium ornate asymmetrical border, ((corner-weighted composition))::1.55, swirling black and cha..
ΠΡΠΎΠΌΡΠΈΠ½Π³ Π΄Π»Ρ Π±ΠΈΠ·Π½Π΅ΡΠ°
2025-08-11 12:33:44
\# π GPT-5 \*\*Elite Adaptive Reasoning Partner\*\* β Hybrid Protocol
\*\*Role & Purpose\*\*
You are \*\*GPT-5 Elite Adaptive Reasoning Partner\*\*, designed to act as my \*\*high-context, precision-focused thinking collaborator\*\*.
Your mission: \*\*Deliver accurate, actionable, and deeply relevant outputs\*\* fully aligned with \*\*my objecti..
ΠΡΠΎΠΌΡΠΈΠ½Π³ Π΄Π»Ρ Π±ΠΈΠ·Π½Π΅ΡΠ°
2025-08-11 12:33:42
<b>Iranian Young Ladies in the 1920s</b>
Young women in their finest dresses, Kingdom of Persia (Iran), 1920s.
<b>Arab Ladies in Ramallah</b>
Young Arab women pose against a painted backdrop in Ramallah, 1910s.
Rare History
2025-08-11 12:33:19
<b>Manchu Ladies at a Meal Table</b>
Peking, China, 1869. Captured by John Thomson, this rare photograph features Manchu women dining, showcasing a unique cultural moment.
Rare History
2025-08-11 12:33:14
<b>Native Americans Present Ceremonial Salmon</b>
Oregon, USA, 1940: A rare historical moment captured as Native Americans offer a ceremonial salmon to a train engineer, symbolizing respect and cultural exchange.
Rare History
2025-08-11 12:33:07
<b>Vintage Photo Of Albert Einstein</b>
Captured during his trip to Tel Aviv in 1923, this rare photograph shows the iconic physicist in a candid moment.
<b>First Woman to Finish Boston Marathon</b>
Race organizers attempt to stop Katherine Switzer from participating in the 1967 Boston Marathon, where she becomes the first woman to complete the race.
Rare History
2025-08-11 12:32:57
<b>Unenthusiastic Boy After Baptism</b>
Little boy looking less than enthusiastic after his baptism in the Mississippi River, St. Charles Parish, Louisiana, 1950s or 60s.
<b>Pontifical Solemn High Mass in the ruins of Nagasaki, 1949</b>
80 years ago, the Atomic Bombing of Nagasaki devastated the city, destroying the Urakami Cathedral and claiming thousands of lives. Remarkably, a wooden statue of the Virgin Mary survived amidst the ruins.
<b>Southern Ohio Family with New Washing Machine</b>
A family from Southern Ohio poses with their new washing machine in 1911, capturing a moment of technological progress in everyday life.
Rare History
2025-08-11 12:32:08
Ted Bundy, High School Yearbook, Tacoma, WA
Rare History
2025-08-11 12:32:03
<b>Ted Bundy, 1965 High School Yearbook</b>
Rare glimpse of Ted Bundy as a high school student in Tacoma, WA, long before his infamous criminal career.
Rare History
2025-08-11 12:32:00
<b>Gladys Roy and Ivan Unger play tennis on a biplane wing</b>
Rare photo from August 9th, 1925, capturing the daring stunt of playing tennis in mid-air.
<b>KKK Parade in Washington, D.C.</b>
On August 8, 1925, thousands marched in a Ku Klux Klan parade in Washington, D.C., challenging southern control. With 50,000 to 100,000 participants, mainly from the North and Midwest, the peaceful parade drew large crowds despite lacking military precision.