Trying Jasper Art For the First Time - Text to Images

Introduction [00:00]

In this video, we take a first look at Jasper Art, a brand new beta feature that allows users to generate images from text prompts. Although not yet publicly available, you can join the waitlist using the link provided in the description. Throughout the video, we explore various prompts and analyze the results, showcasing the impressive capabilities of this AI-powered art generation tool. Despite some minor imperfections, Jasper Art demonstrates remarkable potential, generating high-quality illustrations and artwork in a matter of seconds. As we delve into the future of AI-generated content, this video offers an exciting glimpse into what Jasper Art can achieve.

Sumo Wrestler [00:39]

The first prompt we test is “a sumo wrestler eating a taco while working at a computer.” The results are astounding:

• The generated images capture the essence of the prompt, depicting a sumo wrestler enjoying a taco while sitting at a computer.
• While some images appear slightly unrealistic, others are incredibly lifelike, with the sumo wrestler’s facial expressions conveying genuine enjoyment.
• The attention to detail is impressive, with additional elements like a secondary computer in the foreground adding depth to the scene.
• Although the food occasionally appears inconsistent with the background style, the overall composition is remarkably well-executed.

The speed at which Jasper Art generates these images is equally impressive, taking only about five seconds to produce the output.

Apple Logo [01:34]

Next, we attempt to generate a thumbnail for a hypothetical YouTube video reviewing an Apple product. The prompt is “a man holding an Apple computer logo over his head.” The initial results are promising but not quite perfect:

• The images include all the requested elements: a man holding an Apple logo above his head.
• However, the compositions appear slightly unsettling, with the man’s facial features and expressions appearing somewhat unnatural.
• Modifying the prompt to “a man triumphantly and excitedly holding an Apple logo over his head” yields similar results, with the man’s teeth and eyes still appearing slightly off-putting.

It seems that the more photorealistic compositions tend to have a slightly dystopian feel, whereas the illustrated artwork generally produces more satisfactory results.

Dragons and Warriors [02:48]

Using Jasper’s prompt suggestion feature, we generate images based on the prompt “painting of two dragons fighting a warrior in a castle.” The results are impressive:

• The paintings depict two dragons engaged in combat, although the warrior is not always clearly visible.
• The artwork is of high quality, resembling illustrations one might find in a graphic novel.
• While not all outputs perfectly match the prompt, the overall quality and style of the paintings are remarkable.

Cartoon Dog [03:44]

We then test a more basic prompt: “a cartoon dog running into the street and getting a bone.” The results vary in style and quality:

• Some images feature a misshapen dog, but the overall composition remains interesting.
• The more lifelike depictions of animals tend to have a slightly unsettling appearance.
• Paintings and artwork generally produce better results compared to the more realistic renderings.
• In some cases, the background appears stylized like a painting, while the cartoon dog is juxtaposed in a more realistic manner.

Robot [04:38]

In an attempt to generate a more complex scene, we prompt Jasper Art with “a computer becoming sentient and taking a paintbrush from a human.” The results are somewhat mixed:

• The images are creepy and weird, with some elements appearing difficult to interpret.
• Simplifying the prompt to “a robot taking a paintbrush from a human” yields better results.
• The robot designs are visually appealing, with one particularly cool-looking robot standing out.
• However, the human hand in some images resembles a skeleton, adding to the dystopian feel.

Human [05:20]

To refine the previous prompt, we change it to “taking a brush away from a human,” emphasizing the word “away.” The results improve:

• The images better capture the intended action of the robot taking the brush away from the human.
• Some robots appear more humanoid or resemble the character Gumby, adding an interesting stylistic twist.
• One image depicts a humanoid figure seemingly trapped inside a robot’s soul, showcasing the creative potential of the AI.

Man waving [05:51]

For the final prompt, we request “an illustration of a man waving goodbye.” The resulting artwork is reminiscent of van Gogh’s style, with distinctive lines and patterns.

Key Takeaways

• Jasper Art is a powerful AI-driven image generation tool that can create high-quality illustrations and artwork from text prompts in a matter of seconds.
• While some generated images may appear slightly unrealistic or unsettling, particularly when aiming for photorealistic compositions, the overall quality and creativity of the output are impressive.
• Paintings and illustrated artwork tend to produce more satisfactory results compared to the more lifelike renderings.
• Prompt specificity and word choice play a crucial role in guiding the AI to generate the desired output.
• As an add-on to the existing Jasper account, Jasper Art is priced at $20 per month per user and is currently available through a waitlist.
• The technology behind Jasper Art hints at a fascinating and exciting future for AI-generated content, with potential applications in various creative fields.

In conclusion, Jasper Art demonstrates remarkable potential as an AI-powered image generation tool, capable of creating high-quality illustrations and artwork from text prompts. Although some outputs may require refinement, the overall quality and speed of the generated images are impressive. As the technology continues to evolve, tools like Jasper Art are set to revolutionize the way we approach creative tasks and content creation.