By now it is virtually certain: artificial intelligence is the technology that will most change the way we live and work in the near future. Since the launch of ChatGPT-the artificial intelligence that seems to be able to answer our every question-interest in this and other examples of artificial intelligence is exploding.
Industry experts and investors are already stretching their ears, but users, companies, and the merely curious have also begun to follow artificial intelligence-based app trends with interest.
Artificial intelligence platforms and apps don’t just aim to keep us company and answer our disparate questions, but will most likely help us in many other ways: they will improve the experience of online searches, support businesses in customer care, enable the creation and optimization of articles, illustrations, and videos, suggest programming codes, and offer advice on how to increase productivity in business cycles.
But what are the artificial intelligence-based apps currently on the market? What is the technology that makes them work? And what exactly do they do? Let’s find out details and news about the products that have already been released and those that are about to be launched.
Here are all the artificial intelligence examples, apps and platforms to watch in the coming months.
- ChatGPT – currently the world’s most popular chatbot, based on conversational artificial intelligence
- Google Bard – Google’s alternative to ChatGPT, designed to revolutionize the world of search engines
- Midjourney– an example of artificial intelligence for creating images from a text command
- DALL·E 2 – another text-to-image artificial intelligence app to watch closely
- Writesonic – artificial intelligence technology applied to the creation of articles and content for social media
- Synthesia – an example of artificial intelligence for creating presentation videos
- GitHub Copilot – an app that helps developers write software source codes using artificial intelligence
ChatGPT
Where else could we start but here? ChatGPT is the artificial intelligence app that has attracted the most comments and interest in recent months. The blogpost with which it was launched in November 2022 is already considered by many to be a historic event for technology, on par with the introduction of the first iPhone in 2007.
Technically, ChatGPT is a chatbot, or conversational artificial intelligence app. The app understands natural language and consistently answers questions posed by users using the same language. ChatGPT can be asked a little bit of everything and in a completely natural way, such as, “Can you translate this text for me?” or “What is the French Revolution?” Among the requests made daily by the more than 600 million users who use it each month many are about poems, programming codes, essays on certain topics, high school assignments, or philosophical conversations.
ChatGPT is currently based on the GPT-3.5 model, an artificial neural network that has been trained through hundreds of billions of online texts. The artificial intelligence app was developed by OpenAI, a company founded by Elon Musk and Sam Altman, with substantial investment from Microsoft.
Of course, ChatGPT, at the time of writing [March 2023], also has limitations. In fact, the training database stops at 2021: this means that ChatGPT ignores what happened in the last two years. Peculiar isn’t it? Moreover, it is a text-only artificial intelligence: it can neither understand nor generate images, videos, voice commands. Although some of these barriers will probably be overcome with the next upgrade to GPT-4.
Currently the software can be accessed for free online from here, after registration. Unfortunately, due to the high volume of requests, it is often difficult to use it.
In summary, ChatGPT is:
- An app based on conversational artificial intelligence (chatbot).
- It reacts to textual input by providing articulate responses in natural language
- It is text-based only; it does not include audio or images
- It is currently free
Google Bard
Google Bard is announced as Google’s answer to ChatGPT. Currently [March 2023] this artificial intelligence app has not yet been made public and is only accessible to a very small number of users.
Examples of artificial intelligence: Google Bar
For now we know that Google Bard, like ChatGPT, is a conversational artificial intelligence app, more commonly referred to as a chatbot. Although exact details about its capabilities are not yet available, Google Bard should be able to consistently respond to different kinds of input from users in the form of conversation. Specifically, Google is working on integrating Google Bard with its search engine so as to revolutionize the experience of our Internet searches, making them more reliable and closer to our language. Similarly, ChatGPT is in the process of being integrated with Bing, Google’s main competing search engine.
Google Bard is built on LaMDA (Language Model for Dialogue Applications) technology, a family of artificial neural networks applied to language and developed by Google starting in 2020.
In summary, Google Bard is:
- A chatbot similar to ChatGPT, developed by Google
- It was announced in February 2023, but is not yet public
- It will improve the Google search experience
Midjourney
Simplifying a lot, Midjourney is the image version of ChatGPT.
Midjourney isan artificial intelligence app of the AI text-to-image type, that is, artificial intelligences that create images from a user’s textual input. The user, for example, can ask the software to create an illustration of a woman on a bicycle in the style of Picasso or an alpine landscape with cyberpunk elements.
After initiating the first prompt (that’s what the user’s textual input is called in the jargon), the app returns four images ready to be reworked further until the desired result is achieved: for example, one can ask the software to remove elements such as buildings in the background, to use a particular format, or to apply a totally different style.
Midjourney was launched in its beta version (which remains the currently available version) in July 2022 by a small independent lab based in San Francisco. The artificial intelligence is based on machine learning and has been trained on an endless database of digital images.
Already in its early months, this artificial intelligence-based app has attracted interest and criticism. The June 2022 cover of the Economist featured Midjourney, while a careful debate on legal, ethical and philosophical issues around the use of this type of artificial intelligence has been led by digital artist LRNZ and some colleagues.
You can activate a free trial of Midjourney or use it with a paid license. You can access Midjourney from here.
In summary, Midjourney is:
- An artificial intelligence app that creates images from text
- It allows you to rework images by adding details, styles or technical parameters
- It is still in beta but usable with paid plans or a free version
DALL-E 2
Like Midjourney, DALL-E 2 is an artificial intelligence-based app for creating text-to-image type images.
Introduced in 2021 by OpenAi, the same company that owns ChatGPT, the app is currently in its second version, which has already been enhanced and upgraded from the basic software. The artificial neural network that makes it work has been trained on a large database of images paired with description text.
DALL-E 2 thus exploits the same technology as ChatGPT with the difference that it interprets natural language and generates corresponding images. For example, by entering the command “a gorilla playing computer games,” the app generates a realistic image. The app can be used not only to create a specific image from scratch but also to modify it through simple text commands: if we no longer like the gorilla as a character we can ask the app, “replace the gorilla with a young teenager.”
At present, Midjourney and DALL-E 2 are the two main examples of text-to-image type artificial intelligence: so it will be useful and interesting to follow their development to see if they will differ. For now, the main differences between the two apps are mainly noticeable in style, with DALL-E 2 tending to return more photographic and realistic images than Midjourney.
DALL-E 2 can be accessed from here. You can create your first images for free and buy additional credits to increase its capacity.
In summary, DALL-E 2 is:
- The benchmark artificial intelligence app for text-to-image, along with Midjourney.
- It tends to have a more realistic and photographic style than its competitors
- It was launched by OpenAi, the company that owns ChatGPT
Writesonic
Writesonic is an artificial intelligence-based app for creating and optimizing texts and articles-a kind of digital copywriter that leverages an AI engine.
Using Natural Language Processing (NLP) models and machine learning, Writesonic aims to create more or less complex articles, content for social media, and text for online advertisement. The app also takes advantage of artificial intelligence to improve texts that have already been written, for example, optimizing search words or helping to manage images to be attached to the article.
The first version of Writesonic, still quite rudimentary, was launched in 2020 by its creator, Indian computer scientist Garg Samanyou, with the idea of combining the potential of artificial intelligence with human creativity. Meanwhile, the company Writesonic has also made its own alternative toChatGPT: the chatbot Chatsonic.
Like ChatGPT, ChatSonic is a conversational artificial intelligence-based app that responds to text inputs. Compared to its more famous competitor, ChatSonic, however, can also respond by generating images and understands voice orders as well. The app also has been trained on a database that has been updated to the present day, as opposed to ChatGPT’s database which is stuck in 2021.
You can try Writesonic for free within a certain word limit, after which you need to activate paid plans. The app is accessed here.
In summary, Writersonic is:
- An artificial intelligence-based app for creating and optimizing articles.
- It is used to create or optimize articles, blogposts, social media content and advertisements
- The same company also offers ChatSonic, an alternative chatbot to ChatGPT
Synthesia
Synthesia is a video creation platform that relies on artificial intelligence. The app was founded in 2017 by a group of researchers from Stanford, UCL, TUM, and Cambridge universities.
Synthesia works very simply via a browser: by entering textual input, it creates presentation videos (at the moment, decidedly static) in which an avatar recites the text provided by the user. The advantage for companies is that they can make many videos without having to involve an entire video team and employ expensive instrumentation such as microphones and cameras.
Behind the app’s operation is an artificial intelligence that adapts speech to the movement of the avatars’ lips and synthesizes natural-like voices (here is an example of a promo video for a financial app).
Currently in Synthesia you can choose from over 85 avatars based on real actors and 120 languages (including, of course, Italian). In some languages you can even vary the accent. To access a demo or contract a quote suitable for your needs, you need to contact the company. The basic version costs €26 per month.
In summary, Synthesia is:
- An artificial intelligence-based app for creating promo videos.
- Artificial intelligence synthesizes natural voices and synchronizes speech with the lips of avatars
- To request a demo version, it is necessary to contact the company
GitHub Copilot
Did you know that artificial intelligence-based apps can help developers themselves in their work? GitHub Copilot, for example, is a platform for supporting developers in the creation of programming code-a sort of programmer’s helper.
When compiling software, the app suggests to the user different ways in which the code can be completed in order to speed up and streamline the programmer’s work. GitHub Copilot is based on GPT-3 technology, the same technology behind ChatGPT: in this case, however, the artificial intelligence model is specifically designed to understand and generate both human and programming language.
At the moment the app does not seem to go beyond the role of a simple assistant and does not seem to be able to take care of a complex system on its own. Will it succeed in the future?
You can try GitHub Copilot here.
In summary, GitHub Copilot is:
- An artificial intelligence-based assistant for compiling programming code.
- Artificial intelligence understands and generates both human language and source codes
- The app currently does not seem to be able to deal with complex systems on its own.
And have you tried any of these AI-based apps yet? What do you think of these early examples?