AI paints art, solves maths, codes and completes novel tasks without direct training ... but yet not lifting EBITDA
Dr. Mozhgan Tavakolifard
Author of The AI Awakening | Founder at AI Alchemy Hub | TEDx Speaker | Tech Entrepreneur & Startup Advisor | Executive Board Member
Will we have soon a visual equivalent of books? Something like Audible, though in a visual format? If so, I would get less bored training at gym ??. The image above is a high resolution photograph of an oil slick on a puddle, on a city sidewalk after a rainstorm, reflecting the skyscrapers above. It is created by DALL-E2 from OpenAI that came in beta a week ago! Since then, I cannot stop watching what one million users who got access started creating and sharing on Twitter. DALL-E2 is an AI system that creates realistic images and art from a description in natural language.
If we set aside all legal and ethical challenges this can bring on, the potential and impact on industry would be enormous even without quantum computing.
领英推荐
Imagine how this can revolutionise design, art, education, content creation ... or how this can impact development of digital twins, for example, in combination with knowledge graphs and platforms like Metaverse. We are not far from to see the impact on som industries like fast fashion, media and advertising, manufacturing,construction, ..., set aside entertainment and movie industry.
Earlier OpenAI researchers also created LEAN, a language model that could solve two problems from Elite International Math Olympiads and Codex that could translate natural language commands into codes. Similarly, Google DeepMind unveiled AlphaCode, an AI model that can write computer codes in a range of different languages. It actually participated in a coding contest and came in within top 28%. Google researchers have achieved the remarkable result of training a robot that completed 24 (out of 28) novel tasks in novel settings without any additional training!!!
It is difficult to say what this all mean for HR transformation and workforce. Although, I think it is clear that we cannot treat AI as a cool thing in our Innovation labs creating "pilots" anymore. We have to figure out how to create scalable solutions in operation that truly impact the business. We have to start treating AI as a Business Case with proper funding mechanism and value tracking. We have to speed up with all missing pieces such as responsible AI, governance, op-model, way of working, platforms, talents, culture and adoption (oh this last one can eat AI strategy for breakfast ??) ... We asked more than 1500 of our clients in Accenture across 16 industries. Three in four executives fear they risk going out of business in five years if they don't scale with?AI!
Global Sales Officer at Accenture | Unlocking Potential, Purpose and Value for Businesses | C-Suite Advisor | Manages Multi Billion Dollar Business | Champion of D&I and Sustainability
2 年Very interesting Mozhgan, a recommended read for all