Indistinguishable from Magic

Economics
There has been a lot of talk about large language models (LLMs) vs natural language processing (NLP) since OpenAI unveiled ChatGPT. The product represents a technological advancement beyond what the consensus expected

As we've talked to different clients about what it might mean for them, we're realizing this quote is true:

Arthur C. Clark

Many of our clients are still figuring out what ChatGPT means for their businesses in terms of threats and opportunities. But more than anything, there's a bit of awe in its capabilities. One CEO told me "I feel like I'm in the future". We're at the beginning stages of adopting this technology. It's clear there will be many use cases and figuring them out quickly will be important. Here's how we're thinking about it and how we think it relates to the broader macro environment:

Back to basics: the value chain

The best way to understand how this technological advancement might change the way we all do business is to take a look at the value chain model below:

Four Week MBA

The framework above lists out all the activities a business uses to optimize its competitive moat.

- Inbound logistics is how a business sources materials/labor

- Operations is the step where those materials/labor are turned into products/services

- Outbound logistics speaks to delivery of products/services to customers

- Finally, marketing and sales activities are all about communicating & selling the product

- The service triangle at the end represents customer support

- The 'support activities' above 'primary activities' refer to activities like HR, Tech, and Procurement that make the primary activities viable

When I look at this model I see three areas in a company's value chain that LLMs and NLPs have the potential to change:

It Will Be Easier to Outsource Operations, Outbound Logistics, and Services

Language is still one of the largest barriers to outsourcing effectively. Miscommunication and poor expectation setting can result in logistical nightmares. Common issues that arise from this type of miscommunication are quality issues in the product/service, out of scope work, and internal team frustration.

The ability to translate from one language to the other at native fluency is something Google, Apple and other tech companies have been struggling to get right. ChatGPT has solved it. Integrations with companies like Missive (a shared email management platform) make it easy to hire an entire team that speaks little of a native language, and still function at a high level. This will put continued downward pressure on domestic wages. We should also see the secular macro trend of globalization accelerate in speed as adoption of this technology moves up market into larger companies.

Automation & Doing More with Less In Marketing, Sales, and Engineering

Chat GPT can already autogenerate content in marketing that needs to be lightly edited and checked for errors. The technology can also act as a code assistant, generating code templates for engineers to work off of. It can write the outline of stories and screenplays. Additionally, it can be used to generate effective subject lines for outbound sales and recommend an entire sales sequence of emails if asked.

What does this all mean?

To us, it means that businesses will be able to use these functions to get increased productivity without adding headcount. The trend is well underway already, but we think it will quicken. Take a look at the graph below:

Source: Bank of America

Easier for New Entrants to Reduce Competitive Moats

Some of the hardest business problems to solve fall into the category of matching problems. Netflix and Spotify have built entire businesses around solving these types of problems: how do you find a movie for a person wants when they want to watch it?

These problems will get much easier with these LLM advancements. Take a look below:

ChatGPT is not trained in real-time. The latest year it has information on is 2021, currently. But it won't stay like that for long. Already we're seeing companies like Microsoft begin to embed it into their services with fascinating results . Going forward, we think barriers to entry will fall for D2C businesses that rely on recommendations to match supply and demand.

What it means for you

1. Outsourcing has never been easier. Look for ways to harness this new tech and cut your total costs. Language barriers are quickly becoming small hurdles.

2. You can use your current staff and this new automation to achieve growth without hiring more people.

3. In fact, this technology can boost a company's overall capacity and productivity by a great deal -- have your engineering team evaluate whether helpful code can be generated. It's worth considering how much more you can accomplish more with less HC by adopting this tech.

4. Embed ChatGPT into your marketing department. Let it generate content, then make some light edits or enhancements. It's much faster than paying Fiverr or another platform to do the same job.

5. Ask ChatGPT to recommend five "effective outbound sales subject lines" and use the one that you like the best -- anecdotally, we've seen high open and click rates when we send outbound email materials using this methodology.

Final Thoughts:

There was another quote I wanted to include and couldn't find space for it elsewhere -- Marc Andreessen said this:

“Any new technology tends to go through a 25-year adoption cycle.”

This is the very beginning of ChatGPT's adoption cycle. If taken seriously now, I think you'll be able to deliver value to your clients faster with fewer people.

Feel free to reach out with feedback or if you just want to chat about the market conditions or technology.

e: rob@bobsbookkeepers.com

e: matt@bobsbookkeepers.com

Thanks everyone,

Rob & Matt

Please read last month’s letter here – FRAME 1: WATCHING CYCLES REVERSE

Key takeaways from last month were:

For Profitable Companies:

1. M&A should be on the docket, valuations have plummeted, now is the right time to buy

2. Accelerate hires you’ve postponed

3. Be judicious w/ digital marketing spend, YoY growth is slowing in credit card GMV

4. Bank loans for profitable companies will be more available at these interest rate levels

For Unprofitable Companies:

1. Manage your business to 16-18 months of runway

2. Pull on your venture debt

4. Outsource to fill gaps

4. Do price increases to extend runway (if elasticity of your customer permits)

For the TLDR go to the bottom of this email and read the “What does it mean for you” section.

About Bob's Bookkeepers (or shameless plug):

​Bob's Bookkeepers is a team of financial experts providing bookkeeping and strategic financial guidance for startups and later stage companies. We have a deep understanding that no startup's needs are the same. That is why we offer tailored services to navigate the complexities of sustaining and growing your company. We leverage our 20 years of working with startups to provide insights and guide founders. We use technology to assist companies in better managing their finances to sustain and grow. You can contact us here.

New?

If you're new to this content and want to subscribe, please use this link