AI: a list of tools you can use right now
Whether for business or leasure, AI offers an unrivalled range of tools suitable also for beginners
Original article by Caleb Kraft on Make Magazine
With AI popping up in the news daily, you may be wondering how you can actually play with it. There are many ways you can take advantage of the latest and greatest AI tools, in pretty much any medium you want — audio, video, text, images, and more. And not just for creating specific art or writings, but even for editing and other workflow enhancements.
A brief understanding of AI tools
Before we talk about the various types of Artificial Intelligence tools available in the market, let us first give a brief understanding of AI tools.
It is clear that in the years to come, Artificial Intelligence is expected to assist in all major developments and completely change the way things are done today. It is likely to be an inseparable helping hand in all the major industries. At the moment, businesses need to understand the probable benefits of AI that they can have on their business operations. The need for such tools arises giving humans the freedom to make decisions and let a machine play other predefined roles. These tools not only help you develop but also make significant contributions to optimizing networks and workflows.
Here’s a list of tools you can use right now to experiment with the most cutting-edge computer brains available.
All the rage right now. We’re constantly seeing new features, capabilities, and methods for creating imagery from only a text prompt. Late last year DALL•E 2 was announced and people were blown away. Google’s Deep Dream and some style transfer stuff had been out there to play with, but generally speaking the public wasn’t using prompts to create images from scratch. The tech had obvious errors and downsides — each image was different, consistency was an issue, hands and faces were distorted. Now, there are several different methods and new features every day that push the boundaries of what we thought possible. You can even create a character and use them as a base to create consistent imagery in various styles and situations.
Debate around the legality and morality of this is heated. Programmers are using the artwork of famous artists as training material for these systems, and as templates for their outcome. Some artists feel this is theft of their work.
No matter where you fall on that argument, you probably agree that this feels like an inflection point, much like the invention of digital photo editing tools like Photoshop that changed the industry forever.
[The featured image on this post was created in DALL•E 2 with the prompt: “A person making a red robot using AI technology. Insanely high quality, insanely intricate detail, studio photo, masterpiece, sharp focus, 8k artistic photography, ominous matte painting, dramatic studio lighting, trending on cgsociety, vibrant, epic, intricate, trending on artstation, volumetric cinematic perfect lighting.”]
Some image software you might already be using is integrating AI features. For example, Photoshop is rolling out things like automatic sky replacement, subject selection, and even the ability to change the season within a picture. If you’re using Blender, there’s a free plug-in that will use AI to generate texture files from prompts. And Canva, the free online design suite, has added AI image generation as well.
Another area that’s seeing insane growth and technological achievement is text generation. These systems can create marketing text, carry on conversations, and even write simple stories, essays, reports — a threat to college professors and search engines alike.
Written by Chat-GPT
PRO: Chat-GPT is an innovative tool that allows makers and DIY enthusiasts to access a wealth of knowledge and information in a convenient and easy-to-use manner. With Chat-GPT, users can ask questions and receive detailed, accurate responses in real-time, enabling them to troubleshoot problems, learn new techniques, and get inspiration for their projects.
One of the best things about Chat-GPT is that it is constantly learning and improving, which means that it is able to provide more and more accurate and helpful information to its users over time. …
Overall, Chat-GPT is an invaluable resource for makers and DIY enthusiasts, providing them with quick and easy access to a wealth of knowledge and expertise. Whether you are a beginner or an experienced maker, Chat-GPT is sure to become an essential part of your toolkit.
CON: Chat-GPT is an absolute disaster for makers and DIY enthusiasts. Far from being the helpful and convenient tool that it claims to be, Chat-GPT is nothing more than a time-wasting, error-prone mess.
The responses provided by Chat-GPT are often completely inaccurate, providing users with completely wrong or useless information. This not only wastes the user’s time, but can also be dangerous, as relying on incorrect information can lead to serious accidents or injuries.
Overall, Chat-GPT … should be avoided at all costs by makers and DIY enthusiasts. There are far better resources available that actually provide accurate and reliable information, rather than the garbage provided by Chat-GPT.
AI is really getting into video production with many tools to aid in difficult editing tasks. You can have AI remove backgrounds, capture motion, rotoscope, and even paint out objects automatically.
Free software like Blackmagic Design’s DaVinci video editor are also getting AI tools like the ability to isolate vocals in a noisy environment. This could drastically improve videos shot by beginners and small YouTube channels that don’t have the gear or experience to get a good mix.
The world of spoken audio production has not been spared. You can go from creating narration from scratch to editing pre-recorded clips with the help of AI. Our very own Dale Dougherty uses Descript to edit the Make:cast podcast!
Descript: AI for Audio (and Video) Written by Dale Dougherty (co-founder of Make:magazine)
I am a word person. I like to work in words, sentences, and paragraphs. I also like to talk to people and record conversations, and I started doing that during Covid to create Make:cast, our podcast. The problem was editing sound files. I tried the open-source Audacity and Adobe’s Audition, but it was difficult to adjust to editing timelines. Then I found a new tool, Descript, that did something I thought was game-changing.
In Descript, I upload audio files produced from recording a conversation on Zoom. Then I generate a transcript of the conversation. Now the amazing part: I can edit like a word person, moving, deleting, or rearranging text in one window, and both the audio and the video will automatically reflect the changes I made in text. In other words, editing audio and video became more like editing a document rather than a timeline.
Descript also has a few other tricks based on AI. It can identify and remove filler words, like “uh” and “ah” and “you know,” in a way that’s mostly undetectable to the ear. It also allows you to create a “synthetic” version of your own voice, which requires your reading or speaking for about 20 minutes to train the AI. It is spooky to hear a voice that sounds so similar to your own, speaking words that you type. This artificial version of your voice is not quite good enough because it fails to get intonation right, but it can be a useful tool to insert a last-minute correction.
Descript released a completely new version in November called Descript Storyboard. The goal is to make video editing as easy as editing documents. This new release changed much of the interface, including keyboard shortcuts that created a challenge for its longtime users. It added the idea of breaking up a video into scenes where you can easily add other elements such as graphics and sound. While it does have a timeline view, most of the editing is done within the document itself. For those who find the learning curve for typical video editors like Adobe Premiere too steep a climb, Descript could make video production more like creating PowerPoint slides.
With the new release — and an investment by OpenAI Startup Fund — Descript represents a shake-up in content creation as well as demonstrating how AI tools can be embedded to introduce new features.
With these tools you’ll save time in both the creation of music and the editing process. Simple repetitive tasks like isolating vocals can be offloaded to an AI. Some can even create full musical compositions simply by flipping a few switches.
Code is text — but with more restrictions. For example, if a paragraph of prose has some peculiar wording, you can typically figure out what the meaning is. Code, however, breaks down if there are errors. These systems are hoping to create AI code helpers and even write entire chunks of code using AI.
As with image generation AI, there is a lot of discussion about whether these systems are infringing on people’s work. An AI could, in theory, choose to include a block of code you wrote verbatim, without your explicit permission, claiming it as its own work.
- GitHub Copilot
Beware the Mechanical Turks
Some folks, determined to catch the hype train, are launching services advertised as AI that are possibly just outsourced to cheap labor. Try doing a Google search for the product you’re looking at to see if people are discussing its legitimacy — especially for complex services like 3D model creation and video production from text.