Generative AI — What it is, why we care (a lot), and why you should too
Written by: Jakob Friberg, Data Analyst & Michael Rotstein, VP Business Development
AI and Machine Learning aren’t new; they’ve evolved over decades. But every now and then, technology reaches a new breakthrough, which is what we’ve witnessed over the past few months. Generative AI services like DALL-E, MidJourney, and ChatGPT can create new content like text, images, audio, video, and code. They are good at it – and will get better, fast.
This is a big leap in technology which will change how creators create content and how we build digital products. It’ll change what we, as users, expect from the digital products and services we use. It will most likely change society as a whole, just like the internet and modern smartphones did.
The latest progress in AI has incited fierce competition between tech giants to become the next platform on which a new generation of services will be built. Microsoft invested in OpenAI, the company behind DALL-E and ChatGPT, and Google and Meta were not far behind in announcing their own capabilities and investments in similar technology.
How to think about Generative AI
From a technology perspective, Generative AI encompasses a group of machine learning models capable of generating content based on the input you provide. For example, it can write a blog post on a specific subject, summarize an article, write boilerplate code, or generate a photo-realistic image in a specific style for your presentation.
One of the technological breakthroughs came from a Google paper published in 2017 called “Attention Is All You Need”. This paper presented a new architecture for natural language processing, where one of the advantages of the new architecture was that it could be trained on unstructured data — meaning the model could make use of the world’s largest source of unstructured text; the internet.
This makes it possible for the new generation of Large Language Models (LLMs) to perform logic reasoning, answer questions, translate sentences, and more without having been explicitly trained to do so. Technology aside, what truly made a difference in the public’s eye was how OpenAI packaged their GPT-3.5 language model into an accessible, easy-to-use product, ChatGPT. As a result, ChatGPT reached 100 million users in just two months. And for comparison, it took TikTok – one of the largest existing online services – 9 months to get that many users.
Currently, the internet is exploding with new products and services built on this new generation of language models, from Big Tech and startups alike. Already yesterday, OpenAI released their new generation of the large language model, called GPT-4. GPT-4 is multimodal which means it will be able to combine and mix media (text, images etc), further expanding the possibilities of what you might create with your brand.
In our upcoming posts, we’ll explore some of these new products, and what kind of tasks they help users solve. In the meantime, we’ve listed a few products below that you can explore immediately to get a feel for what has already been achieved and to help spark your imagination of what the future might bring.
Some of the most popular AI tools out right now for you to check out
1. ChatGPT
A chat-bot built on OpenAIs large language models, designed to respond to text-based queries and generate natural language responses.
2. Stable diffusion
An open sourced text-to-image model capable of generating photo-realistic images given text input.
3. Midjourney
A text-to-image application integrated as a bot into Discord, capable of generating images from text input. Recently updated to the v5 version, which vastly improves the application fidelity.
4. Notion AI
An AI-powered tool created to improve productivity and workflow in Notion through the Claude LLM from Anthropic.
5. Bing search (requires waitlist signup)
Microsofts new type of search engine incorporating OpenAIs large language models to augment search results.
Bontouch has worked with AI and Machine Learning for more than seven years. We’re running labs exploring technology and use cases with our partners, incorporating machine learning into products, building chatbots, voice assistants, and more. With this blog post, we will start to share our thoughts and explorations on Generative AI and the extensive field of AI.
If you want to know more or explore what Generative AI means for you and your brand, don’t hesitate to reach out to us at [email protected] so that we can explore this together.