AI as Sustaining Innovation
Where's the money?
Adobe's Firefly: A Glimpse of AI's Future in Creative Industries
On Tuesday, Adobe announced "Firefly," a generative AI model tailored for creatives or those who use Adobe products daily. Firefly not only generates images, but also incorporates AI-driven photo editing techniques that are scarce in other AI text-to-image tools:
Highlight parts of an image to generate variations
Type prompts to edit specific parts of images
Generate custom styles and textures from sketches
Edit videos with text
While these tasks can be accomplished with existing tools, they are rarely unified within a single platform, let alone one that consistently produces high-quality initial images. Will Adobe be the first to achieve this feat?
Microsoft's Copilot: AI Integration Across the Board
Despite my three AI-focused posts this week, I haven't mentioned Microsoft's "Copilot for everything" announcement. While numerous AI tools generate text, integrate with email, create presentations, or transcribe text, Copilot stands out by integrating with Microsoft Office and Teams. This means users never have to leave Outlook, Word, Excel, PowerPoint, or Teams to access AI's value. The AI even understands the context of meetings and emails, streamlining PowerPoint presentation creation and Outlook sharing. In light of this, would anyone gamble on a new-to-market AI-powered word processor?
Google has responded with their own Copilot-like tool (unnamed as of yet, but integrated with Google Workspace). The functionality mirrors Copilot, reinforcing the trend of tech giants infusing AI into existing products to enhance them.
Features vs. Companies
BensBites, a daily email summarizing AI news and newly launched AI-driven companies, has become a favorite of mine. Reviewing these lists repeatedly makes it evident that many of these "companies" are merely "features" that will eventually be incorporated into a leading product. Here's a selection from yesterday's list, accompanied by my tongue-in-cheek descriptions:
Tripnotes: Travel-centric ChatGPT
Bob from Magrathea AI: ChatGPT "assistant"
Chatblade: ChatGPT Command Line Interface
Vowel: AI tool for meeting summaries
Chatshape: DIY AI Chatbots
You-tldr: YouTube video TLDRs
Magic-flow: No-code AI integration for apps
Pair: AI-assisted coding tool
An upscaling tool for 3D ultrasounds
OpenTheo: Bible-focused ChatGPT
A URL scraping Slack bot
SQL tool for auto-loading ChatGPT
Watermelon: GPT-powered code interpreter
Zing: Bing clone
Sustainability Goals AI: Company sustainability goal creation via ChatGPT
Lex: AI-powered word processor with "writing feedback"
Halfway words: ChatGPT for word unions (e.g., crunchy + soft = firm; truth + lie = ambiguity)
While many of these tools may offer improvements over using ChatGPT directly, most seem easily replicable by existing products. For example, how long until YouTube adds a "summarize" feature? Microsoft Copilot already summarizes Teams meetings, is Zoom far behind? Lex generated buzz last year, but can it compete with Microsoft Word and Google Docs?
Opportunities for the Little Guys
That's not to say all of AI's potential will be monopolized by established companies. Innovative startups can still disrupt complacent incumbents, particularly in industries like education, healthcare, and legal services. I recently spoke with a lawyer friend who admitted, "I charge by the hour. If I use AI to work faster, I need to sell more. It breaks my business model." He has an incentive not to innovate – until someone else redefines his industry. He has an incentive not to innovate - at least until someone else comes along and innovates his business model out of existence.
There are a few companies I am already aware of that are using AI to radically change industries. I will cover some examples in a future post. Maybe in a week that is closer to 80% marketing and 20% AI…
Keep it simple,
Edward