The Best AI Inventions of 2023

Spotlighting the Pioneering AI Models, Products, Projects, and Tools Shaping the Future

TIME recently unveiled its list of the best inventions of 2023, highlighting several groundbreaking AI innovations. For our readers' convenience, we've compiled and summarized these AI achievements. They are categorized into AI models, products, projects, and tools. Let's explore these remarkable inventions!

AI models

  • OpenAI GPT-4: OpenAI’s premier system is designed for generating, editing, and iterating on creative and technical writing tasks. It excels in activities such as composing songs, crafting screenplays, or adapting to a user's writing style. Dive into our discussion of the GPT-4 here and explore our research on the OpenAI company here.

  • OpenAI Dall-E 3: This text-to-image model is adept at generating visuals from textual descriptions. Built natively on ChatGPT, it enhances prompt generation. Read our coverage of generative AI models, including diffusion-based models that are at the core of Dall-E 3.

  • Runway Gen-2: This multimodal AI system is equipped to produce innovative videos using text, images, or video clips. Discover more about the Runway company in our profile.

  • Meta SeamlessM4T: A foundational multilingual and multitask model that seamlessly translates and transcribes across speech and text.

  • Stability AI Stable Audio: Employing a latent diffusion model architecture, this system generates audio based on text metadata, audio file duration, and start time.

Products

  • Humane Ai Pin: This standalone device is paired with a software platform that leverages AI. Designed as a wearable for clothing, it integrates a suite of sensors, facilitating ambient and contextual interactions.

  • DedroneCityWide by Dedrone: This platform enables organizations to monitor drone activities in their airspace without the hassle of hardware installation or maintenance. Users can designate alert zones and instantly get notifications when drones encroach on these specified areas.

  • TrailGuard AI: Tailored for national parks, this security system identifies, intervenes, and aids in apprehending poachers. Beyond this, it also enhances intelligence on illicit poaching networks, assisting authorities in combatting illegal wildlife trade.

Projects

  • Project Gutenberg Open Audiobook Collection: A collaboration between Microsoft, MIT, and Google transformed the entire Project Gutenberg Collection into a vast library of over 800 audiobooks. Utilizing state-of-the-art AI for text-to-audio conversion, they achieved exceptional voice-acting quality.

  • ALERTCalifornia and Cal Fire: ALERTCalifornia employs cutting-edge technology, facilitating data-centric decisions to prepare for, respond to, and recuperate from natural disasters. As a public safety initiative, its mission is to grasp the intricacies of wildfires and other natural threats, assess their immediate and prolonged impacts on humans and ecosystems, and guide management determinations.

  • So-VITS-SVC: This project contains an open-source code that allows the training of a neural network using a singer's voice, subsequently generating any song with AI-derived vocals. Enthusiasts frequently exchange voice models of various artists, from Bad Bunny and Taylor Swift to James Hatfield, within Discord communities.

Tools

  • Photoshop Generative AI: With features like Generative Fill and Generative Expand, users can seamlessly add, delete, or augment content within images directly in Adobe Photoshop. These functionalities are enhanced by the Adobe Firefly generative AI, which operates through straightforward text commands.

  • Alitheon FeaturePrint: This avant-garde Optical-AI offers items an identifier analogous to human fingerprints — a singular and enduring marker distinguishing one in millions. By serializing items, FeaturePrint affords capabilities such as track & trace, counteraction against counterfeits and gray markets, and the curtailment of human inaccuracies. All these benefits are harnessed merely through a photograph taken with a conventional camera.

  • Audioshake: AudioShake can split any audio into separate instrument parts, even if the original doesn't have separate tracks. This is great for remixing, syncing to video, and more. It can also single out dialogue, music, and sound effects, useful for film, dubbing, and voice synthesis. Plus, it can quickly transcribe and match lyrics to a song using only the original audio.

Every day, we post helpful lists and bite-sized explanations on our X (Twitter). Please join us there:

Subscribe to keep reading

This content is free, but you must be subscribed to Turing Post to continue reading.

Already a subscriber?Sign In.Not now

Join the conversation

or to participate.