AI: The Pace of Change is Mind Blowing

What happened?

Open AI (Chat GPT) held their first developer conference, and it produced some major changes. I will quickly summarize what they announced and recommend to those that are curious to watch the replay or Open AI’s summary. Arrowside Capital does not own a position in Open AI.

Sam Altman, CEO of Open AI, announced the following improvements:

• GPT-4 Turbo: An improved version of GPT-4 and it will come in two versions, one for text only and one for text and images.
• GPT-4 Turbo is more capable and has knowledge of world events up to April 2023. It has a context window of 128k, so it can fit the equivalent of 300 pages of text in a single prompt.
• Sam announced that the company reduced prices across the platform to pass on savings to developers. GPT-4 Turbo are now 3X cheaper than GPT4 and output tokens are 2X cheaper.
• Users can build their own GPTs: Users can build their own GPTs for productivity or other uses. Developers can use new sources of data to have the model analyze.
• GPT Store for user created bots: This will come later this month and initially be for only registered creators. Sam alluded to paying creators of the most popular ones.
• New API for developers to build assistants: This would allow developers to build their own agent-like applications on top of the model. For more on the future of agents, see Bill Gates’ post that can be found here. https://www.gatesnotes.com/AI-agents

• DALL-E 3 API: DALL-E is a text to image model and now has API capability.
• Defend against copyright claims: Sam promised to have Open AI defend against copyright claims that stem from the use of the company’s products.

Takeaway: Be very excited and get on board!

The pace of change here is staggering. Many pundits are intimating that the improvements in the last month exceed that of the prior year. Judging from the comments from Sam Altman and others, we should get used to this rapid pace of improvement.

There was a lot to get excited about as a user. Most importantly, Open AI is treating their LLM as an ecosystem, Apple style. This means that many developers will be able to benefit from developing GPTs on top of Chat GPT. This will create a lot of new processes that will enable users to take advantage of the power of the model, which keeps getting better. I personally cannot wait to see what people come up with.

One other thing to note, and it may be the most important. The costs coming down for developers is a big deal. One of the knocks on the very broad adoption of Gen AI models is the cost per inference, industry jargon for a query. By reducing the cost to developers, the company will drive further innovation, and for this particular technology, the sky is the limit.

AI added two to my team without making a single hire….

Speaking of GPT development, I went down the rabbit hole and tried it for myself. I played around with creating my own GPTs for processes that I would need a person to do or where I wanted to try and save some time. I created a GPT to edit my writing (yes, this was edited..). I also created a GPT that summarizes company earnings conference calls. I was very specific with my instructions and found the results to be really accurate for those that I compared to my own takeaways. I even included instructions to give me a summary about the tone of the management team during the call, whether they were optimistic or not. I found this to be pretty

spot on as well, and I plan to use this tool alongside my normal process. For those that are interested in some ideas for writing these, I share my GPT at the end of this post.

One thing that struck me was how GPT like process improvements will continue to raise the bar for everyone. At some point soon, roles on a team that are geared heavily toward reporting facts or gathering information will be obsolete. To me, this isn’t scary, rather it is going to make work product much more useful, if an individual or team has skill.

AI: Initial Thoughts on Market Implications

I have been thinking a lot about this topic lately. It is always a good idea to be on the right side of change. This is no exception, in fact, the stakes are really high right now due to the pace of change that is underway. (see above)

I have gone back to the vault in my memory to pull out learnings and experiences from both the PC Era and the Internet Era, both of which I had a front row seat for as a PM.

Here are my initial thoughts, subject to updates:

• Expect Wider Dispersion of Returns: I believe that we have already started a period of wider dispersion in returns for individual companies and their stocks. We have been in a long period of lower dispersion following the Global Financial Crisis. There were a number of factors that led to that, but I believe that the pace of AI will return us to a period more like the 1990s. This will increase single security risk, as well as reward…
• Because the Magnitude of the Change and the Pace are so Rapid: I believe that this will take many by surprise, as will the overall pace of innovation ushered in by AI. There will be companies that will be slow to adapt or will be shut out from the benefits of AI. The returns for these companies will be much worse than the market, and especially relative to the companies that are on the right side of this change. For the beneficiaries, the spoils will far exceed the average return. Alpha comes back in a big way.

• Good Public Market Opportunities: I believe that public markets will have great opportunities for taking advantage of AI, and it will extend beyond the technology sector. The public markets will lead out in terms of more immediate benefit for investors, and that has already started to happen with the separation of the Magnificent 7 this year to date.
• Melt Up? Maybe, and maybe not for everyone: I think that we could see a period of a melt up of the market as the enthusiasm for AI takes hold. The enthusiasm is there but as soon as companies report revenue and expense benefits due to the use of AI, the market could see a significant bump as speculation creeps in. I think it is more likely that this happens for certain sectors and industries rather than the market as a whole. More to come on this in the coming weeks.

Corporate Prudence – It’s Back, but Will It Stay?

It is hard to argue that we just went through a very unusual period from 2019 to today. The pandemic wreaked havoc on the economy, got a zero interest rate policy response, and the fallout touched all parts of the economy. In 2020 and 2021, we witnessed a different investment cycle, one that had tons of FOMO and many companies adopted a growth at all costs mentality. There have been tons of carnage for companies, both public and private, that had tried to grow at unsustainable rates and hired way too many people. It also exposed many companies in the technology and consumer sectors that had essentially been buying growth by spending wildly on ads to drive traffic.

What will the next several years hold for the companies that showed good corporate prudence, keeping debt to a reasonable level and not spending to try and grow at unsustainable levels? So far this year, we have witnessed better results and price returns, at least in the strategies that we manage. But, if we go back to interest rates in the 2.5-3% range, does that behavior come back? Will the money sources, such as VC and aggressive growth investors reward revenue growth only, without regard for profitability?

My personal belief is that this would seem to be a no brainer that prudence should remain in vogue. However, I see some signs that the behavior before the Great Rate Hikes is still around to some degree. Some VC firms are throwing money at AI startups with just an idea. Open AI is doing multiple rounds of share sales, both primary and secondary, with the valuation escalating each time. Granted, this is a pretty special company, but there is a degree of speculation and hype that can be unhealthy. We may not be there yet, but it has been a bit eye opening to see some of this behavior creep back in so soon.