Artificial intelligence (AI) isn’t new to the news publishing industry – it has been used for several years to create template articles about predictable, data-centric news stories like public company quarterly earnings, updated government economic data, and sports scores. The Washington Post has its Heliograf tool, Bloomberg has Cyborg, Yahoo! and the AP have used tools from Automated Insights, United Robots has built tools for local news publications across Sweden, etc.
Generative AI tools have reached a much higher caliber now, both in the data being input into the models (billions of websites) and the sophistication of the output. The current wave of new tech trains on vast amounts of existing content in order to output new content based on it, as opposed to extracting structured data from trusted sources to publish new articles that feature and describe that data within a structured template.
This is the start of a new paradigm shift in publishing, media, and the broader digital economy. There are still many imperfections but the pace of improvement is accelerating. This paradigm shift will be at least as big for news publishers as the shift to digital. It will necessitate a different way of producing content and running an organization. Some publications will disappear as they fail to keep up and new ones will rise that are native to this new technology environment.
Below are some generative AI tools that news publishing companies might find helpful. I will update this from time to time on this page. Email me with suggestions of what other generative AI tools to add (or with reviews of what did vs. didn’t work well for you). Also reach out if you lead a news publisher and want to participate in some group discussions I have organized about this.
Last update: 29 March 2023
Elicit: finding relevant research papers, summarizing those research papers, and pulling info from across several papers to compare what they say about a specific data point or topic.
Text Creation (Drafting Articles/Posts, Summarizing, Copywriting)
ChatGPT: the driving of much of this innovation and the tech most of these other tools are built on top of.
Claude by Anthropic: heavily funded OpenAI competitor focused on chatbot needs for businesses, with emphasis on avoiding inappropriate content
Lex: writing assistant for newsletters, blogs, etc (built on GPT)
GrammarlyGO: AI enhancement to Grammarly’s existing editing suggestions, helping you write more concisely / articulately by rephrasing things
Jenni: grammar, plagarism check, autocomplete, and suggested citations for writers
Hypotenuse: website copy, blog posts, & social posts
Lately: social media posts
Writer: AI for content marketing copy that fits your detailed brand guidelines & makes sure colleagues keep their writing on-brand; maybe could be used by a news publisher to make sure writers stick to the style guide or for drafting branded content posts.
Hexo: platform to help you create your own AI image generator using Stable Diffusion
Let’s Enhance: image enhancer & upscaling for making images larger regardless of original resolution
Podcasting: Transcription (Speech-to-Text) & Editing
Whisper: OpenAI’s API for audio-to-text transcriptions. Works for many languages and can translate to English, but most of its training data was in English so accuracy in other languages is less.
Descript: transcribes podcasts and lets you edit the audio file using the transcript (like a Google Doc); can auto-remove filler words
Adobe Podcast: background noise cancellation, other editing tools
krisp: background noise cancellation for VOIP calls (including podcast recordings in Zencastr and elsewhere)
Recut: removes overly long silent pauses from podcast & video files
+ Mubert: AI generated, royalty-free music for podcasts
Audio Editions of Articles (Text-to-Speech)
Video Creation & Editing
RunwayML: image & video creation and editing suite; generate images and video (from text or input photo/video), expand images beyond what was in the original, erase items from videos, make any video clip slo-mo, etc.
Fliki: text to video with AI voiceover
Steve AI: text-to-video for product, podcast, etc. videos (turns blog post into video); using library of stock content and AI generated imagery
Veed: simplified video creation and editing for people without technical editing expertise
Vidyo: creating social media ready clips from long videos
Zubtitle: subtitling videos
Neural Space: speech to text, text to speech, language recognition, & translation for local languages across Africa, Middle East, and Asia
Resemble AI: voice cloning with speech-to-speech and language translation
Synthesia: Platform for lip syncing of videos so mouths of people in them match the new text; focused on localizing business videos across 65 languages
Perspective: content moderation tool (API) by Google’s Jigsaw using AI to detect toxic comments
Vectara: large language model (LLM)-based search tool for finding content within websites and apps
There are many other tools for businesses in general (accounting, legal, taxes, software development, design, etc.) that are relevant to news organizations also but I won’t try to list them here for now.