Issue #40. Businesses are using more AI

First, in case you noticed, I’m sorry about the wrong subject line in the last email. I forgot to update it from a previous newsletter.

On today’s quest:

— Using EditGPT with Word
— Business use of AI is up
— Consumers turned off by AI
— Big AI copyright case moves forward
— Word watch: AI nationalism
— AI costs for developers coming down
— Press 1 to speak to a human

Using EditGPT with Word

I’ve recommended EditGPT before — a neat tool that lets you see tracked changes on text you put into ChatGPT for edits — but many of you are newly subscribed since then, and hey, a refresher never hurts anyway. This week, Marcella Fecteau Weiner has a nice tutorial on using EditGPT with Word.

Businesses are using more AI

Business spending on AI doubled year over year according to the payment processing company Ramp. OpenAI and Anthropic topped the list of vendors that businesses transacted with for the first time on Ramp cards from January to June 2024.

The top AI vendors on Ramp in Q2, based on overall customer count, were OpenAI, Anthropic, Midjourney, and Grammarly, which has been touting its AI foundation and features recently. The mean for company spending on AI in Q2 was about $100K.

According to the company, spending patterns and stickiness reveal that companies are increasingly convinced of AI’s value. Businesses also increased their spending on independent contractors. This makes me wonder if companies are replacing employees with a combination of AI and contractors.— The AI Daily Brief podcast 

Consumers turned off by AI

A new study found that consumers were far less likely to buy a hypothetical smart TV if the description included “artificial intelligence,” and the reaction extended to other products as well. "When AI is mentioned, it tends to lower emotional trust, which in turn decreases purchase intentions," said lead author and Washington State University clinical assistant professor of marketing Mesut Cicek in a statement. — Futurism

Big AI copyright case moves forward

A judge in California allowed a class action copyright infringement lawsuit to move forward against the AI art companies Stability AI, Midjourney, Runway AI, and DeviantArt. Although not all the artists’ claims survived, the vibe is that they think it’s a big win. — Ars Technica

Word watch: AI nationalism

Concerns that AI is so powerful that whoever controls it has huge national security advantages is leading to “AI Nationalism” — countries trying to control the technology and limit other countries’ access to it.

For example, Nvidia, the dominant AI chip company, is based in California and isn’t allowed to sell its chips to China because of export restrictions. But some in China have found ways around the rules and have acquired chips anyway both through loopholes and underground trades.

AI costs for developers are coming down

A lot of the news over the last couple of months has been about LLM models getting faster and cheaper — not something that’s terribly exciting for us to know, but I’ve seen it so much I thought I’d mention it so you can feel in the know.

An example is this announcement from Anthropic which says it’s new Claude prompt caching reduces costs “by up to 90% and latency by up to 85% for long prompts.”

In the long run, more efficient models should lead to better products. For example, Anthropic says prompt caching will help users “Talk to books, papers, documentation, podcast transcripts, and other long-form content.” It will “bring any knowledge base alive by embedding the entire document(s) into the prompt, and letting users ask it questions.”

Press 1 to speak to a human

A White House press release touting “a new effort to crack down on everyday headaches and hassles” talked about requiring companies to “let customers talk to a human by pressing a single button.”

Although this story isn’t directly about AI, it caught my attention because I believe call centers are one of the top industries adopting AI.

Without AI, it seems like the new rules would make companies hire more customer service reps as we all learn we can “press 1” to avoid endless, frustrating phone trees. I’m sure companies are hoping AI representatives will be good enough that we won’t want to.

Quick Hits

Top 5 misconceptions editors have about AI (1. AI is not safe to use with clients' text.) — Erin Servais

8 predictions for what generative AI means for the business of editing (1. AI will never understand the author’s voice)— Intelligent Editing

What is AI sidequest?

Using AI isn’t my main job, and it probably isn’t yours either. I’m Mignon Fogarty, and Grammar Girl is my main gig, but I haven’t seen a technology this transformative since the development of the internet, and I want to learn about it. I bet you do too.

So here we are! Sidequesting together.

If you like the newsletter, please share it with a friend.

^{Written by a human.}