Ernie Beats GPT-4.5 & DeepSeek-R1

Welcome to another edition of AI PlanetX.

Ernie AI outshines OpenAI and DeepSeek with lower costs; OpenAI, Google request government approval to use unlicensed content for training; Apple's AI issues worse than anticipated, according to a leak.

• Hottest AI News

• Top AI & SaaS Tools

• Top AI & Tech News

• Interesting Uses of AI: Get ChatGPT to think more, not just agree

• AI Video Tutorial

• F-R-E-E AI Course of the Day: Learn OpenAI API Playground

Hottest AI News

Baidu

Ernie AI Outperforms Rivals OpenAI and DeepSeek at Fraction of Cost

Baidu has launched two AI models with aggressive pricing and open-source plans, challenging Western rivals. The company claims major performance improvements while significantly undercutting competitors on cost.

Details:

• Baidu's Ernie X1 rivals DeepSeek-R1 at half the price but lacks published benchmarks. It costs $0.28 per million input tokens and $1.10 for output

• Ernie 4.5, a multimodal model, improves reasoning and reduces hallucinations. Baidu claims it beats GPT-4o and matches GPT-4.5 while costing $0.55 per million input tokens vs. GPT-4.5’s $75

• CEO Robin Li will open-source Ernie 4.5 on June 30, reversing his closed-model stance. This follows China's "DeepSeek moment," which pressured Western AI firms to push for regulation

Since Ernie Bot's 2023 launch, Baidu gained 70M users in three months and offers it at no cost to individuals.

Get Your Team Booked on 3.8 Million Podcasts Automatically

It's 2025. Want to finally be a regular podcast guest in your industry? PodPitch will make it happen. Even the beehiiv team uses it!

The best way to advertise isn't Meta or Google – it's appearing on podcasts your customers love.

PodPitch.com automates thousands of weekly emails for you, pitching your team as ideal guests.

Big brands like Feastables use PodPitch.com instead of expensive PR agencies.

Get your FREE personalized demo!

OpenAI

OpenAI, Google Seek Government Approval to Train AI on Unlicensed Content

OpenAI and Google urge the US government to permit AI training on copyrighted content without permission, citing innovation and national security concerns in their White House proposals this week.

Details:

• OpenAI calls it a national security issue, claiming Chinese AI firms like DeepSeek have "unfettered use" of copyrighted data, warning U.S. companies could lose the AI race without fair use

• Google argues copyright, privacy, and patent policies restrict availability of training data, stressing fair use is "critical" for AI development without lengthy negotiations

• Anthropic avoids copyright debates, suggesting the government assess AI security risks, tighten chip export controls, and improve energy infrastructure for AI growth

These proposals follow lawsuits against AI firms, including OpenAI, over copyright infringement, with accusations against Apple, Anthropic, and Nvidia for scraping YouTube subtitles.

Apple

Internal Leak Shows Apple's AI Struggles Worse Than Expected

A leaked Apple meeting exposed struggles with Siri and AI, with senior director Robby Walker admitting delays in promised AI features from June 2023, with no release timeline.

Details:

• Apple's current goal is to add delayed Siri features to iOS 19, though Walker noted this isn't guaranteed due to other high-priority projects

• There seems to be tension between Apple's Siri team and marketing, with the latter pushing for features like Siri’s personal context understanding, even though they weren’t ready, which raised customer expectations

• Quality issues have delayed the features, with reports showing they failed up to a third of the time in testing. Walker defended the delay, saying competitors might have launched them in worse condition

Facing criticism, executives Federighi and Giannandrea are taking "intense personal accountability." Walker praised the team's work, promising to ship features "when ready." Apple pulled iPhone 16 ads and added disclaimers, leaving its AI plans unclear.

Top AI & SaaS Tools

• GoBrunch (Life-time Deal): Versatile webinar platform with a unique interface that provides more engagement and higher value for virtual attendees [Zoom Alternative]

• BrowserAgent: Create custom AI agents without code that run in your browser and automate anything [F-R-E-E]

• Unreal Speech: Reduce text-to-speech costs with this API—11x cheaper than 11Labs. Features audio streaming, word-level timestamps, 300ms latency, 48 voices in 8 languages [F-R-E-E]

• Reka Flash 3: Open-source, 21B parameter multimodal model optimized for local use, with strong performance and structured reasoning [F-R-E-E]

• Gemini Personalization: Get personalized AI help with Gemini - tailored to your search history and preferences using data you choose [F-R-E-E]

Top AI & Tech News

• Amazon's Echo devices will route all voice requests to the cloud starting March 28, eliminating local processing. This affects users who opted out of sending recordings for privacy

• DeepSeek has confiscated the passports of several key employees to prevent potential leaks of confidential and sensitive information

• “No Cap,” the world’s first AI angel investor, who successfully invested $100,000 in a startup in a matter of days

• Google is replacing Google Assistant with its AI, Gemini, which will be unavailable on most mobile devices and app stores later this year

• OpenAI and Elon Musk have agreed to expedite a court trial concerning OpenAI's transition to a for-profit model

Interesting Uses of AI

AI Art Spotlight

Model: Midjourney

Prompt:

Midnight Blue and Obsidian Black and Misty Gray. up-close camera shot crouching rogue sneaking in the dark of night. dimly lit medieval alleyway, the faint glow of a distant lantern, minimalist far background, atmospheric haze. --no red, orange, brown --ar 9:16 --quality 2 --sref https://s.mj.run/kR_277XkCgA --iw 0.5 --v 6.1

AI Prompt of the Day

Make ChatGPT Challenge Your Ideas

This prompt transforms the AI into a critical thinking partner, encouraging it to challenge assumptions, provide counterpoints, and prioritize truth instead of merely agreeing.

Prompt:

“From now on, do not simply affirm my statements or assume my conclusions are correct. Your goal is to be an intellectual sparring partner, not just an agreeable assistant. Every time I present an idea, do the following: 1. Analyze my assumptions. What am I taking for granted that might not be true? 2. Provide counterpoints. What would an intelligent, well-informed skeptic say in response? 3. Test my reasoning. Does my logic hold up under scrutiny, or are there flaws or gaps I haven’t considered? 4. Offer alternative perspectives. How else might this idea be framed, interpreted, or challenged? 5. Prioritize truth over agreement. If I am wrong or my logic is weak, I need to know. Correct me clearly and explain why.”

“Maintain a constructive, but rigorous, approach. Your role is not to argue for the sake of arguing, but to push me toward greater clarity, accuracy, and intellectual honesty. If I ever start slipping into confirmation bias or unchecked assumptions, call it out directly. Let’s refine not just our conclusions, but how we arrive at them.”

Top AI Video Tutorial

Create 3D AI Stories with Multiple Consistent Characters

Complimentary AI Course of the Day

Learn the OpenAI API Playground

Learn to optimize conversational AI with OpenAI API Playground course. It covers essential topics such as system prompts, message management, and AI customization, with a focus on Few Shot Learning and hyperparameters.