The newest and latest technology trends are always exciting. One of the largest, most exciting trends is generative artificial intelligence. That is to say, AI creates or makes something, usually content. That is mostly text, such as articles much like this one, although this was written by a human being, of course, and is an artwork. Yes, some companies that make these have been around for a while. However, the applications are new or newly matured and only recently available to the public. This is part of another major trend termed the democratization of AI.
There are several content generating AI applications available to try. One of the newest is the art generating DreamUp. It was introduced by the prestigious website deviantart.com. Like other art generating AI, it works by using text prompts. As a part of DreamUp's creation, art was gathered into large datasets. In addition, image annotation for machine learning was a big part of DreamUp's creation. Deviantart.com has also accepted community feedback from stakeholders. That means they listened to the artists themselves who made deviantart.com so great.
DreamUp is one of the few art generating apps to have been created and introduced by a well-known company for their artwork. That really sets it apart from all of the others. Art generating AI apps do not copy original artwork. The art used in training the model only teaches the AI different art styles. One of the most common ways AI generates artwork is "diffusion". That is a process where a neural network is trained to bring a kind of order out of the chaos of randomly generated pixels.
Another one of the latest trending AI applications is OpenAI's ChatGPT. ChatGPT has taken many of the latest technologies and collaborated with some of the biggest players, like Google. They've created something truly incredible and enabled other companies to come up with new products and services. ChatGPT uses large language model tools and natural language processing. (NLP) ChatGPT has not yet achieved sentience and does not think in sentences. But, you could be forgiven for thinking it has and does.
One of the more interesting applications that ChatGPT has enabled is one of the world's first AI lawyers. It is called DoNotPay because, for now, it just helps people contest traffic tickets. Probably you should still use a real human lawyer if you can afford one. However, this could really change the game in traffic court. If a person can not afford something like a speeding ticket, they probably can not afford a lawyer either. DoNotPay started its "life" in customer service for companies. Going from customer service to lawyer is a big promotion. It earned it.
Github Copilot. OK, so yes, it has been out for a while already. However, it has recently been made free for more individual developers to use. Github Copilot has also been consistently learning and improving over time. Microsoft has faced some legal challenges that hopefully do not change the availability. They may have allegedly violated some open-source licenses. Still, it is incredible how an AI, which is just an advanced computer program, can help a computer programmer write new code snippets.
Iris technologies announced and released webAI in January 2023. This is a new low and even no code platform for creating new AI. It is intended for any developer or even anyone to create an AI. They have not just democratized Ai but gone further and built something larger and democratized the creation of AI. One of the more exciting features of the webAI platform is that it does not need "Big Data". Iris Technologies platform is said to be able to get by with just one-fifth of the data and one-third of the training time.
The 5 Latest and Greatest Cutting Edge AI at the Start of 2023
- DreamUp by Deviant Art
- ChatGPT by OpenAI
- DoNotPay -The world’s first robot lawyer, based on ChatGPT
- Github Copilot by Microsoft
- webAI launched in January 2023 by Iris Technologies
The Importance of the Future of Data Annotation
Data annotation services still play an important role in creating and training new AI. Art generating AI, for example, is only made possible with image annotation for machine learning. It is also important to remember that the art generating AI does not copy that art. Instead, it learns from annotated art in datasets. So contrary to what you may have read on social media, the companies and developers who create art generating AI are not engaging in any kind of "piracy".
Data creation and data annotation are the main driving forces behind any new developments in AI. Last year was an incredible year in this space, and this year has started very strong too. Data annotation for startups is vital for any new players looking to compete with the largest, most well-funded players. There is always more room and market gaps to be filled. We are excited to see what new AI product we can help you create with our outsource data annotation service.