We use the following prompt to get the speech text generated by GPT-3.5 and Llama-3-70B-Instruct-GPTQ:
Transform the provided text into a public-speaking format, ensuring the essential content remains unchanged and the adaptation doesn’t overshadow the original message. Focus on enhancing the text’s vividness and interactivity to improve understanding and audience engagement. Incorporate expressions of opinions and emotions related to key events, and subtly include spoken language features to make the delivery feel more natural and fluid, all while keeping the stylistic elements balanced and appropriate for the context.
input:
[input text]
And converted them into videos using tools from synthesia.
video part1:
video part2:
video part1:
video part2:
video part1:
video part2: