- AI Tool Report
- Posts
- 🤖 Anthropic’s new AI chatbot is better than ChatGPT4
🤖 Anthropic’s new AI chatbot is better than ChatGPT4
🥇 AI For Professionals is officially live
Martin Crowley & Liam Lawson
March 05, 2024
JOB BOARD | PRO | PODCAST | TOOLS | ACCREDITATIONS
🔝Tuesday’s top story: Google-backed start-up, Anthropic, releases Claude 3 and claims it’s better than ChatGPT and Gemini.
To celebrate, we’re also releasing our accreditation alongside a podcast with Jason Flaks, the co-founder and CTO of Xembly (one of our module leaders).
🎙️🎙️Give it a listen 🎙️🎙️
🌤️ This Morning on AI Tool Report
🤖 Anthropic’s new AI chatbot is better than ChatGPT4
📝 OpenAI signs letter agreeing to build AI responsibly
💬 New “read aloud” ChatGPT feature
⛔ AMD hits US roadblock in selling AI Chip to China
🎶 Madonna pioneers AI in her concert tour
Read Time: 5 minutes
CHATBOTS
Our Report: Anthropic (founded by ex-OpenAI employees and backed by Google) has released Claude 3—a series of Large Language Models (LLMs)— claiming it outperforms OpenAI’s GPT-4 and Google’s Gemini.
🔑 Key Points:
Unlike Claude 2, Claude 3 responds to text and up to 20 image inputs in a single request (it only produces text responses, due to a lack of Enterprise demand for images).
It processes data quicker and understands more context, meaning it answers more questions (accurately), understands longer instructions, and gives near-instantaneous results.
It has better reasoning, complex task comprehension, and graduate-level knowledge than GPT-4, scoring 50% in tests vs GPT-4’s 36%, demonstrating near-human capability.
It’s been trained (using hardware from Amazon and Google) on non-public internal and third-party datasets and publicly available data collected after August 2023.
🤨 Why you should care: Following Google’s decision to disable its image generator due to historical misinterpretations, Anthropic has stopped Claude 3 from identifying people, showing awareness for potential bias and hallucinations, and respect for ethical and legal implications.
Which is your favorite: Claude, ChatGPT, or Gemini? |
Are you struggling to find excellent resources to learn hands-on AI skills? Looking to put yourself ahead of the competition? Ready to learn, earn & share your AI accreditation?
Over the past 6+ months, we’ve been working tirelessly with 4 AI industry experts—who have raised a combined total of over $100M in AI-specific funding—to bring you the only AI accreditation you’ll ever need, teaching you valuable, tangible AI skills:
How to save 400 hours per year with an AI executive assistant
How to 4x your communication with custom how-to guides
How to build a functional AI application in less than 1 hour
How to do 5x more with ChatGPT in less than 2 hours
Over 14 days, AI industry experts will teach you the core skillset to start demanding 77% more compensation, with a skill set that only 1% possess.
The only AI accreditation you’ll ever need is now live—sign up now!
Taplio helps you get more LinkedIn followers, reach, engagement, and leads with AI 📈
StoryCraft AI is an AI storyteller that adapts to your style
ArtiverseHub is an AI-powered image-generation tool
Emplibot automatically publishes SEO-friendly articles
LogoAi uses AI to create professional brand identities
ETHICS
Following Musk’s lawsuit (claiming OpenAI was putting profits before people), OpenAI has signed a letter, agreeing to build AI responsibly and mitigate risk to the public.
The letter (also signed by Salesforce) was written by VC firm ‘SV Angel’, and highlighted a “collective responsibility” to “maximize AI’s benefits and mitigate risks” to society.
The letter spotlights the impact of AI, stating that “the balance of its good and bad impacts on humans will be shaped through the actions we as humans exercise.”
Although the letter doesn’t include any actions, those who have signed it are showing a commitment to building AI responsibly, with Sam Altman “excited for the spirit ” of it.
Use the Prompt Engineer GPT by AI Tool report to 10x your prompts.
✅ Cold DMs
copy & paste ⬇️
“I'm looking for a cold DM idea that will leverage the authenticity and relatability of my [brand/company] to engage my [ideal customer persona] and persuade them to take [desired action]."
ACCESSIBILITY
OpenAI has released a new feature called “Read Aloud” which allows users to listen to the chatbot's response to their prompt, rather than read it themselves.
The new feature can speak 37 languages, has five different voice options to choose from, and can automatically detect the language of the text it’s reading.
It’s now available to all ChatGPT 3.5 and ChatGPT 4.0 users on both the web version of ChatGPT and the iOS and Android ChatGPT apps, for no extra cost.
This follows the release of the voice chat feature, which allows users to verbally prompt the chatbot (vs typing), with both features marking a significant step forward in accessibility.
EXPORTS
AMD has hit a major US government roadblock in its efforts to sell an advanced AI chip, which was designed and built specifically for the Chinese market.
AMD designed its new chip to have a lower performance than the chips it sells to countries outside of China, to try and meet US export regulations and restrictions.
But US officials have said it’s still too powerful and they must obtain a license from Commerce's Bureau of Industry and Security before they can sell it to China.
This comes after the US planned to halt advanced AI chip shipments to China to stop Beijing from using cutting-edge US technology to strengthen its military operations.
ENTERTAINMENT
When Madonna sings the 1980s hit “La Isla Bonita” on her concert tour, she uses a text-to-video tool to create moving images of swirling, sunset clouds on the arena screens.
They originally created the effect using CGI, but Madonna felt it was too bland, so they used Runway, a text-to-video tool rivaling OpenAI’s Sora (not yet publically available).
This comes after AI video experts say the emerging text-to-video technology could irreversibly change the entertainment industry, putting thousands out of work.
With Nakkara, you don’t have to settle for just a picture; we turn your AI-generated image (e.g. from Midjourney, DALLE, or Stable Diffusion) into something you can hold in your hand. Just upload a single image, and we’ll send you your “thing” in a few days.
Nakkara is perfect for:
Prototyping
Ideation
Personalized gifts
At Nakkara (Gujarati for “solid”), we use only the highest-quality resin printing, with an incredibly thin, 50 micron layer height, for a smooth, nearly injection-molded finish. We’ll even paint your print for you.
Your vision deserves more than just two dimensions.
We read your emails, comments, and poll replies daily.
Hit reply and let us know what you want more of!
What did you think of this edition? |
