What’s the Buzz in the Valley?
My day-to-day is to talk with people in the Valley (without being InTheValley), and I’ve heard big rumors and expectations about Generative AI.
Bay Area News says it could bring talent back to San Francisco and revive the area. Big companies are training and hiring entire teams to use Generative AI tools. Other creative agencies laid off half of the staff and replaced them with AI tools.
Every investment firm confirms that their portfolio companies are either thinking or implementing a Generative AI tool. And, of course, the world is well aware of the actors, and screenwriter strikes around AI tools.
Some are going as far as saying that AI will create video games and movies in the future (I tend to agree.)
What’s Happening Today with AI Tools?
But how about today? The best architect I’ve met while playing with some AI/ML Python libraries said that he felt embarrassed about how easy his implementation was on predictive analysis using these tools.
Those fancy talent marketplaces are proud to have built a multi-score matching algorithm over the years; now, any small startup can implement the same or even better with OpenAI in less than a week. No joke, I’ve done that myself.
We are, without a doubt, during the implementation phase of the Generative AI tool. We are way past the initial big bang of AI with GPT leading the charge and are now in the frenzy phase.
How Can I Use AI in My Business?
Now this is not a technology problem to solve; it is a product problem: how can I implement these Generative AI tools in my business?
Among those early adopters, there is always confusion about this new technology, a question often overlooked because most jump into how they can use it. So let’s start with the definition first.
What is Generative AI?
You can use AI technology to generate text, images, or videos. These seem to be interim steps to the AGI (Artificial General Intelligence), whose big goal is an AI that can perform (outperform) any human being task.
The timeline was: deep neural networks in 2014 led to the transformer network, which led to GPT in 2017 (text generation, including code generation). In 2021, a transformer-based pixel generative model was the first step in image generation.
What Are Today’s Tools?
For text generation, models like ChatGPT (GPT4) are leading the charge. Code development is made more intuitive with Github Copilot, while image creation is being transformed by tools like DALL-E, Midjourney, and Stable Diffusion.
The world of music isn’t left behind with innovations like JukeBox, and video generation is getting a boost with Gen-2 and Make-A-Video.
Check them out:
- Text: ChatGPT (GPT4)
- Code: Github Copilot
- Images: DALL-E, Midjourney, Stable Diffusion
- Music: JukeBox
- Video: Gen-2, Make-A-Video
What About Regulations?
Given the potential to generate fake content and manipulate people and the potential for scams, such as FraudGPT, it is no wonder that the very creators and investors of AI are pledging to watermark AI-generated content.
From a business standpoint, we are looking for a potentially highly-regulated industry that will narrow down or restrict the current use of your product or business operations, so you should pay special attention to how these changes will impact you.
How Will People Perceive AI-Generated Content?
I wonder if people will see AI-generated content as better or worse than human-generated content.
One thing is clear, from an economic point of view, content creation cost has dropped immensely.
So the future holds a lot more content in less time and higher and higher quality content that is not happening right now because everyone is just getting their heads around how to best use these new powerful tools.
Conclusion
Generative AI is no longer a concept of the future; it’s happening now.
The pace of implementation, the vast array of applications, and the challenges of regulation present a thrilling landscape for innovators, business leaders, and consumers alike.
As we embark on this new frontier, understanding the tools and the ethical considerations will be paramount.
Book Recommendation
For those interested in exploring further, I recommend reading “The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake Our World” by Pedro Domingos.
It provides an in-depth exploration of the concepts and future prospects of AI and machine learning, a must-read for anyone invested in this field.
The content in this article has been enhanced with GPT-4. Curious about the original draft?
Title: Generative AI
My day-to-day is to talk with people in the valley (without being InTheValley), and I've header big rumors and expectations about Generative AI.
Bay Area News is saying it could bring talent back to San Francisco and revive the area. Big companies are training and hiring entire teams to use Generative AI tools. Other creative agencies laid off half of the staff and replaced them with AI tools.
Every investment firm confirms that their portfolio companies are either thinking or implementing a Generative AI tool.
And, of course, the world is well aware of the actors, and screenwriter strikes around AI tools.
Some are going as far as saying that AI will create video games and movies in the future. I tend to agree.
But how about today? Today what's going on with AI tools? The best architect I've met while playing with some AI/ML Python libraries said that he felt embarrassed about how easy his implementation was on predictive analysis. And it is true.
Those fancy talent marketplaces are proud to have built a multi-score matching algorithm over the years; now, any small startup can implement the same or even better with OpenAI in less than a week. No joke.
We are, without a doubt, during the implementation phase of the Generative AI tool. We are way past the initial big bang of AI with GPT leading the charge and are now in the frenzy phase.
Now this is not a technology problem to solve; it is a product problem: how can I implement these Generative AI tools in my business?
As with many other new tools, some will adopt the wait-and-see period (usually those that will get behind), some will take small steps (implementing the tools as part of small steps in the business process), and some will think big and be to the extreme (usually the ones that will rip more benefits of it.)
Among those early adopters, there is always confusion about this new technology, a question often overlooked because most jump into how they can use it. So let's start with the definition first.
What is Generative AI?
You can use AI technology to generate text, images, or videos.
These seem to be interim steps to the AGI (Artificial General Intelligence), whose big goal is an AI that can perform (outperform) any human being task.
The progress was deep neural networks in 2014 led to the transformer network, which led to GPT in 2017 (text generation, including code generation). In 2021, a transformer-based pixel generative model was the first step in image generation.
What are today's tools?
Text: ChatGPT (GPT4)
Code: Github Copilot
Images: DALL-E, Midjourney, Stable Diffusion
Music: JukeBox
Video: Gen-2, Make-A-Video
Regulations
Given the potential to generate fake content and manipulate people and the potential for scams (https://www.pcmag.com/news/after-wormgpt-fraudgpt-emerges-to-help-scammers-steal-your-data), it is no wonder that the very creators and investors of AI are pledging to watermark AI-generated content https://www.reuters.com/technology/openai-google-others-pledge-watermark-ai-content-safety-white-house-2023-07-21/
From a business standpoint, we are looking for a potentially highly-regulated industry that will narrow down or restrict the current use of your product or business operations, so you should pay special attention to how these changes will impact you.
People's perception
I wonder if people will see AI-generated content as better or worse than human-generated content. There is always this view of Us vs. Them, where AI will outperform humans, and humans have intuition, imagination, and emotions that can't be replaced. On the other vision, humans AI assists humans or transcends their human condition to be AI-powered.
One thing is clear, from an economic point of view, content creation cost has dropped immensely. That means you can generate more in less time or invest in increasing quality.
So the future holds a lot more content in less time (which is happening right now) and a higher and higher quality content that is not happening right now because everyone is just getting their heads around on how to use these new powerful tools best.
- Generative AI - 09/06/23
- When to Scale Your IT Department - 08/30/23
- Playing with Fire: Tech ROI in High-Stakes Projects - 08/23/23