Hi AI Futurists,

Today, we’re exploring how automatic prompt revisions might be unintentionally clipping the wings of creativity in DALL-E 3, and what that means for the future of AI-driven artistry. 👀

Best,
Lex

Learn AI Strategies & Tools in the
Growth School Workshop

We have partnered with Growth School to bring you a ChatGPT & AI Workshop — usually $199, but free for our readers 🎁

👉 Register here for free – valid for next 24 hours only!

This workshop has been taken by 1 Million people across the globe, who have been able to:

• Build a business that makes $10,000 by just using AI tools

• Make quick & smarter decisions using AI-led data insights

• Write emails, content & more in seconds using AI

• Solve complex problems, research 10x faster & save 16 hours every week

You’ll wish you knew about this FREE AI Training sooner. Save your seat for $0 now — valid for 100 people only.

Highlight of the Day

Can Prompt Tweaks Undermine DALL-E 3's Power?

Recent research from UC Berkeley suggests that automatically rephrasing prompts with a large language model (LLM) might actually hurt the quality of images generated by DALL-E 3.

In an experiment with nearly 1,900 participants, the study found that DALL-E 3’s performance advantage over its predecessor, DALL-E 2, dropped by 58% when prompts were automatically rewritten by an LLM. While DALL-E 3 still outperformed DALL-E 2, the results indicate that AI-assisted prompt rewrites could be limiting users from fully leveraging the model’s advanced capabilities.

The study also sheds light on an emerging trend: as text-to-image models like DALL-E advance, users are adapting by crafting longer, more descriptive prompts. This suggests that despite AI becoming more sophisticated, human input in the form of well-crafted prompts will remain crucial in unlocking the full potential of these tools. The researchers caution that automatic prompt revisions, while useful for safety and moderation, may not always align with users' creative goals.

With AI models continuously evolving, it seems that the art of prompting won’t become obsolete anytime soon. Instead, it will evolve alongside these technologies, shaping how we harness their power in creative and practical applications.

Summary

• DALL-E 3's performance advantage over DALL-E 2 drops by 58% with automatic prompt rewriting.

• Participants in the study naturally adapted their prompts to be longer and more descriptive when using DALL-E 3, indicating that prompting will remain important as AI models evolve.

• The study highlights potential limitations of AI-assisted prompt rewriting in creative and practical applications.

What we think

Automatic prompt revisions might sound handy, but this research shows they can be more of a hindrance than a help, especially when creativity is on the line. It’s like having an overzealous editor who keeps changing your work — sure, they mean well, but sometimes they just don’t get what you’re going for.

Stick with your human intuition.

👉 Read More

Cool AI Tools

1. Topview.ai

   AI video editor that turns your links or media assets into viral videos

2. Zaver

   Find, analyze & outreach creators from Google Sheets

3. Langtrace AI

   Open-source observability tool that helps monitor, evaluate, and improve your LLM apps

Best of the Rest

• Maybe Friend wasn’t crazy for spending $1.8M on a domain after all

• The FCC proposes requiring robocallers to disclose when they’re using AI

• AI learns athletes' steroid profiles to catch doping with minimal data

• Anthropic tests its "next-generation system for AI safety mitigations”

• OpenAI worries people may become emotionally reliant on its new ChatGPT voice mode

And now your moment of AI art Zen

That’s all for today folks!

• If you’re enjoying the newsletter, share with a friend by sending them this link: 👉 https://www.futureblueprint.xyz/subscribe

• Looking for past newsletters? You can find them all here.

• Working on a cool A.I. project that you would like us to write about? Reply to this email with details, we’d love to hear from you!