Okay, the GPT-3 hype seems pretty reasonable

Published March 17, 2021

This morning TechCrunch covered an interesting round for Copy.ai, a startup that employs GPT-3 to help other companies with their writing projects. GPT-3, or Generative Pre-trained Transformer 3, is a piece of AI from the OpenAI group that takes text from the user, and writes a lot more for them.

As part of the process of covering the Copy.ai round, I got caught up in the idea of AI-powered writing. I’ve long been more curious than afraid of automated writing. So when the Copy team described their very positive impressions of the GPT-3 AI writing tool to TechCrunch during an interview, I was intrigued.

To scratch this newly-formed itch, I doodled around this morning with a competitor of sorts to the Copy team , Headlime. And, freaking heck am I am impressed at what folks have managed to build around the GPT-3 technology.

Sure, GPT-3 can add words to a prompt. But the technology can do a lot more than that. The GPT-3-powered Headlime managed to not only write some medium-good stuff for me, but also managed bring in concepts concerning my reporting beat that were in my head but not in the prompts I provided.

I can’t do better than just show you what I mean. So, here’s what happened when I used Headlime for the first time, sans help.

Here’s the first thing that Headlime showed me, a language selector and a request for a description of the post that I wanted to write. I decided to push the system a bit by just telling it about a piece I need to write in light of today’s market action:

Ha ha, I thought, that will kick it in the teeth and I, a biped of intelligent meat wrapped around some calcium sticks, will feel grossly superior to the computer player. I hit go and then realized that I actually had to provide 500 characters of stuff, so I rambled for a bit to fill in required length:

Time for the next step! Hitting the button brought up a list of possible headlines for the post I was helping create, which were honestly not terrible:

Fair enough, yeah? At this point I was starting to become impressed.

I selected the first headline as it was my favorite and moved along. Next came the work to get an intro put together for the post, a process that involved the strenuous work of clicking a button:

Here are the options proffered:

Again, not bad.

What struck me about these are not merely minor variations on each other. They are structured differently, taking various angles on what I was halfway talking about in the 500 characters of bilge I had fed into the system. I was starting to wish that I had given GPT-3 a bit more to work with up top, as it was trying its best after I had clearly not.

Intro selected, I was brought into a CMS of sorts, where our selected bits were included, and your humble servant was asked to do a bit more writing.

I was happy to oblige, only for the system to stop me and offer to take over:

Having precisely no idea what a credit is, or what two of them cost as I was on a free trial of sorts, I hit the “Write for me” button. This is what came out:

Look at how it finished that sentence I started, even after I used em-dashes! The software gets the next sentence backwards, but is right back on the horse afterwards talking about how higher interest rates make exotic investment classes like venture capital less attractive! I was gobsmacked.

I will keep playing with the tech and the various software wrappers that are being built to productize GPT-3. More notes to come. But I wanted to pause and share my initial delight. This is cool. I can’t recall the last time that technology actually shocked me. But, well played GPT-3, you’re amazing.

Writing helper Copy.ai raises $2.9M in a round led by Craft Ventures

Published March 17, 2021

Copy.ai, a startup building AI-powered copywriting tools for business customers, announced a $2.9 million round this morning. The investment was led by Craft Ventures. Other investors took part in the deal, including smaller checks from Li Jin’s newly formed Atelier Ventures and Sequoia.

The startup is notable for a few reasons. First for its model of building in public. I initially heard of the company through its monthly updates that it posts on Twitter. Thanks to that, I can tell you that Copy.ai generated monthly recurring revenue (MRR) of $53,600. That figure, up 46% from January, works out to annual recurring revenue (ARR) of $643,200.

Copy.ai also shares usage numbers, and, humorously, the number of Twitter followers that its founder Paul Yacoubian picked up in the last month.

The startup is also worth watching because it is part of a growing cohort of companies building atop GPT-3, what its progenitor the OpenAI project describes as an “autoregressive language model with 175 billion parameters.” More generally, it’s a piece of AI that can generate words.

Some investors are rather bullish on startups using the technology. Recently on TechCrunch, for example, Madrona’s Matt McIlwain wrote that “the introduction of GPT-3 in 2020 was a tipping point for artificial intelligence” that will lead to “the launch of a thousand new startups and applications.”

4 ways startups will drive GPT-3 adoption in 2021

So far that’s holding up. Not only has Copy.ai managed to find early in-market traction, TechCrunch has covered a number of other startups busy leveraging GPT-3, including OthersideAi, which raised $2.6 million back in November of 2020, and an “AI Dungeon-maker” called Latitude that also employs GPT-3 and raised $3.3 million this February.

But enough about its cohort. Let’s get into how Copy.ai got built.

AI Dungeon-maker Latitude raises $3.3M to build games with ‘infinite’ story possibilities

Origins

Before founding Copy.ai, Yacoubian was an investor and, it seems, a tinkerer. He played with GPT-3 predecessor GPT-2 when it came out, telling TechCrunch in an interview that he discovered that the tool generated lots of “nonsense,” with the occasional “flash of brilliance.” GPT-3 proved even better in his view, providing something akin to a “50x” improvement on the generation that came before it.

Leaning on Twitter as a distribution method — Copy.ai uses Twitter as a distribution channel, hence its reporting on social media metrics — Yacoubian and his co-founder Chris Lu launched a few different draft-projects using GPT-3. Simplify.so did text condensing, a slackbot was built but never made it to the outside world and taglines.ai was put together to help companies come up with slogans.

That last one found early traction, generating around 700 sign-ups in two days. That was enough of a user base, the co-founders decided, to begin monetizing their tool. Then they decided that the initial concept could be extended to other writing use cases, helping people with myriad distinct writing projects. Copy.ai was formed out of that concept.

The product can now generate text for blogs and products and headlines and the like, based on user-provided word inputs.

What’s odd and nearly antithetical to your humble servant as a writer is that Copy.ai doesn’t want to save you word count, per se. Instead, it generates a number of possible text results from which the customer then chooses. Recall the flashes of brilliance that Yacoubian said GPT-2 could generate? GPT-3 is even better, giving users of Copy.ai even better possible text formulations for their needs. And then the human-in-the-loop plays the editor role, choosing which they want the most and, I presume, tweaking from there.

When it was released back in October of 2020, Copy.ai snagged 2,000 sign-ups in its first two days. Then investors started reaching out.

Quitting their day jobs, Copy.ai became a full-time affair. The unorthodox startup also put together an unorthodox round, raising from what Yacoubian described as “as many people as [they] could.” That wound up being 80 people, give or take.

The round was raised as a capped SAFE, the Y Combinator-favored investing instrument that allows startups to accrete capital from external sources without a formal pricing; instead, SAFEs are often “capped” at a maximum valuation. Copy.ai raised its cap as its fundraising process trundled along.

David Sacks, founder of Craft Ventures, told TechCrunch that he thinks that “natural language generation powered by AI is going to change the way that marketing teams write copy,” adding that amongst startups it is “rare to see such strong bottom-up adoption in so short a time.”

I am honestly a bit excited to see what Copy.ai can do, not because I will use its product — it’s not precisely in my wheelhouse — but because I am rather excited about GPT-3 as a technology. And the startup is an in-market experiment regarding AI and writing. Two things I care quite a lot about.

Early Stage is the premier “how-to” event for startup entrepreneurs and investors. You’ll hear firsthand how some of the most successful founders and VCs build their businesses, raise money and manage their portfolios. We’ll cover every aspect of company building: Fundraising, recruiting, sales, product-market fit, PR, marketing and brand building. Each session also has audience participation built-in — there’s ample time included for audience questions and discussion. Use code “TCARTICLE at checkout to get 20% off tickets right here.