The Future of AI: Trends and Predictions by Michael Ngo

Explores an Intelligent Future with Mosaic Venture Lab at TechRoad 2023

We stand at the precipice of an artificial intelligence revolution. TechRoad 2023 brought together the best minds in AI, mobility, and semiconductors to glimpse the future during their groundbreaking conference in Taipei. I attended as an intern at Mosaic Venture Lab, one of the organizers. The event left me certain that Taiwan will lead the way in our impending AI-powered world.

Keynote speaker Michael Ngo, GM of leading AI English tutor app ELSA, captivated the crowd with his talk, “The Future of AI: Trends and Predictions.” He took us on a thrilling tour of AI’s evolution and demonstrated its vast potential to transform learning. His product exemplified the incredible personalization and efficiency that AI enables.

The Origins of Artificial Intelligence

The famous Dartmouth Conference of 1956 coined the term “artificial intelligence.” Yet decades passed before AI attracted mainstream notoriety. The recent explosion of generative AI caught the public’s attention thanks to tools like DALL-E 2 for image generation.

Generative vs Predictive AI

Michael differentiated between two types of AI — generative and predictive. Generative AI produces novel content like images, videos, music or text. He called them “creativity amplifiers” that save time and costs.

Predictive AI analyzes data to make forecasts. It powers recommendation systems and chatbots. While less glamorous than generative AI, predictive powers much of the internet.

The Future of Generative AI

Michael outlined several trends in generative AI:

Multimodal — Generative models combine text, images, audio and video. We’ve seen tools like Unsplash’s Dream pop up that generate AI art from text prompts. In the future, multimodal AI will create immersive experiences personalized to the user.

Democratization — Open source models like Stable Diffusion make generative AI affordable and accessible. As adoption spreads, expect new innovative use cases to emerge.

Hyperpersonalization — Generative AI can eventually produce customized content for every individual user automatically.

New Use Cases — Michael highlighted the promise of AI assistants and tutors. Companies like ELSA can provide efficient, personalized education at scale.

The Democratization of Knowledge through AI

I loved Michael’s vision of leveraging generative AI to spread knowledge. He traced the evolution of foundation models — the backbone algorithms behind generative AI.

Initially only massive tech firms like Google had the resources to develop these complex models. But open source releases put the power into the public’s hands. This fueled an Cambrian explosion of creativity with models fine-tuned to niche applications.

Specialized generative models enabled startups like ELSA to build tailored products impossible just years prior. Democratized access to knowledge helps realize generative AI’s full potential.

ELSA Corp — Your Personal AI English Tutor

As GM of ELSA, Michael Ngo provides English learners personalized lessons via AI tutors. ELSA assesses proficiency across speaking, listening, vocabulary and grammar. It creates customized learning journeys to address the user’s weaknesses.

I was blown away by ELSA’s features:

• AI-graded spoken English with real-time feedback helps improve pronunciation and fluency
• Virtual conversational practice with topics tailored to user interests
• Grammar and vocabulary exercises adapted to the learner’s level
• Simulated TOEFL and IELTS testing powered by AI scoring

ELSA exemplifies hyper-personalized education powered by generative AI. It unlocks a future where knowledge barriers dissolve through intelligent tutoring.

Taiwan Rising — Pioneering an Intelligent Future

Michael left us with an auspicious vision. He foresees a coming proliferation of AI avatars — interactive virtual assistants simulated using generative video, voice and conversations. TechRoad 2023 proved Taiwan poised to lead us into this future thanks to its tech dominance in AI and semiconductors.

With ELSA as evidence, Michael sees education as an ideal testing ground for AI and generative models. By nurturing Taiwan’s next generation of AI talent, the country will pioneer the intelligent world soon to come.

About TechRoad 2023

Mosaic Venture Lab and Orbit Startups proudly present TechRoad 2023, a groundbreaking tech event in December 2023, in collaboration with NSTC and TTA. TechRoad 2023 spotlights Taiwan’s pivotal role in technology and explores the latest in semiconductors, AI, and mobility. Join us for keynotes, panels, and startup pitches, uniting industry leaders and visionaries. Our mission aligns with Taiwan’s vision for technological advancement and economic growth, making significant strides towards shared goals. Taiwan’s tech prominence shines through TechRoad 2023.

The platform is designed to extend exclusive invitations to distinguished professionals from Taiwan, specifically in the fields of semiconductors, automotive technology, and artificial intelligence. Its core objective is to establish a robust and enduring platform that facilitates interactions and encourages the exchange of emerging trends and cutting-edge technologies among these accomplished individuals.

About us — Mosaic Venture Lab

Mosaic Venture Lab, powered by Audi AG and Continental AG, is a product development-focused platform based in Taiwan. We have a strong focus on smart mobility — Mobility Services, Electrification, Human-Centric / New-data Driven Services, Sustainability, etc.

Text/ Jim Wu
Photo/ Mosaic Venture Lab, Orbit Startups
The cover is Photo by D koi on Unsplash