Intelligent Software, Smartphone Killer, Top LLMs

Greetings, fellow AI explorers! ๐Ÿ‘‹ Welcome to another action-packed edition covering the latest earth-shaking developments in artificial intelligence. Grab some popcorn, and let's dive in.

15th Edition

Could AI transform computing into something unrecognizably intelligent? ๐Ÿค–

Bill Gates makes a bold prediction - within five years, AI could power adaptive "agent" software that deeply understands users, communicating naturally to assist them like a close friend across diverse tasks. 

Rather than making us toggle between disparate apps, agents would provide a unified, conversational interface drawing on comprehensive knowledge of our lives. This advancement could completely revolutionize computing, bringing unprecedented intelligence and personalization.

If agents deliver on their paradigm-shifting potential, we may look back at even today's most innovative software as primitive precursors. Computing could become an extension of ourselves rather than just a passive tool. But this immense power will require thoughtful oversight of privacy and ethics.

AI is about to completely change how you use computers
In 5 years, agents will be able to give health care advice, tutor students, do your shopping, help workers be far more productive, and much more

Pin this: AI wearable devices might free us from phone addiction! ๐Ÿ“Œ How?

After much anticipation, Humane has revealed full details on its pioneering $699 AI Pin. This screenless wearable provides hands-free access to AI through voice commands, gestures, and built-in projection.

Demoed features include ChatGPT and Microsoft model integration, messaging, summarization, translation, visual Q&A, and, remarkably, nutrition analysis via food images taken by the Pin's cameraโ€”humane aims to fundamentally reimagine human-AI interaction by freeing us from phone addiction (watch the demo๐Ÿ‘‡ to be amazed).

As one of the first consumer AI wearables, this launch provides a glimpse into the mainstream integration of ambient computing. The Pin's camera-based applications around health and food also highlight AI's immense potential for personalized wellness.

Who is leading the fierce LLM race after ChatGPT's launch? ๐Ÿ 

With hundreds now vying for pole position, supremacy in large language models remains up for grabs after ChatGPT ignited the generative AI frenzy. From scrappy startups to research giants, all push the boundaries of generative AI through unprecedented computing investments. The LLM race is still in its opening laps. But which are the most capable ones? To find out, follow this list of the best large language models in 2023.

LLM By License Access via Worth Mentioning
GPT-4 OpenAI Proprietary ChatGPT Plus - showcases skills that exhibit human-level performance,
- first multimodal model that can accept both texts and images as input,
- addresses hallucination and improved factuality by a mile
GPT-3.5 OpenAI Proprietary ChatGPT - incredibly fast model and generates a complete response within seconds,
- hallucinates a lot and spews false information frequently
PaLM 2 Google Proprietary Bard - a multilingual model and can understand idioms, riddles, and nuanced texts from different languages,
- very quick to respond and offers three responses at once
Claude Anthropic Proprietary Claude Chat - can take close to 75,000 words in a single window as input
Falcon TII, UAE Open Source Falcon Demo - has outranked all the open-source models released so far,
- can be used for commercial purposes
LLaMA Meta Open Source On request - has been released for research
- Llama 2 is available under a permissive commercial license, whereas Llama 1 was limited to non-commercial use.

That concludes this week's expanded AI updates! Join me next time as we continue surfing the exhilarating wave of artificial intelligence advancement. Who knows what world-changing innovations await around the corner โ‰๏ธ