Meta, post-Q1 2024

Meta is down 15% pre-market on Apr 25, 2024 after reporting first quarter earnings and their outlook for the future. Here is the transcript. Meta reported 4.71 earnings per share, well above the consensus estimate of 4.3. Revenue was in line with expectations too. But, for Q2, they expect revenue between 36.5B to 39B but analysts were expecting 38.5B which means the “midpoint of the range is a disappointment to the Street”. If I am going to be honest, it feels pedantic. In general, investors are “unhappy about profligate spending on AI”. Here’s my thoughts on why I feel differently, and feel happy about it.

Mark said this in the earnings call:

I think it’s worth calling out that we’ve historically seen a lot of volatility in our stock during this phase of our product playbook ‐‐ where we’re investing in scaling a new product but aren’t yet monetizing it. We saw this with Reels, Stories, as News Feed transitioned to mobile and more. And I also expect to see a multi‐year investment cycle before we’ve fully scaled Meta AI, business AIs, and more into the profitable services I expect as well. Historically, investing to build these new scaled experiences in our apps has been a very good long term investment for us and for investors who have stuck with us. And the initial signs are quite positive here too. But building the leading AI will also be a larger undertaking than the other experiences we’ve added to our apps, and this is likely going to take several years.

• Meta is not alone in this bet on AI. Literally every big tech company in software or the semiconductor space is heavily invested here. And I think the public (i.e., non-technical people) believe that “AI” refers exclusively to large language models and generative AI. This is not true. It is one (very promising) technology that scientists and engineers in this space work on, and right now, there is a lot of effort to push the envelope here. But, no doubt, there are other new technologies being developed too. LLMs may not be the exact right architecture (who knows), but investment into AI/ML technology is the right approach in my estimation.

• Meta consistently puts out some of the best research and AI products. And, they have always been heavily invested in this area. Their LLM llama3 is quite good, and Mark says they are doing it as “open source for the dev community and it is now going to be powering Meta AI. There’s a lot that I’m sure we’ll get into around Llama-3, but I think the bottom line on this is that we think now that Meta AI is the most intelligent, freely-available AI assistant that people can use.”

• They are open source. I love that. PyTorch is a great example. It is the go-to ML library, over Google’s Tensorflow. Meta is committed to prioritizing the developer experience which is critical. I remember reading a story, I now can’t find the link, of Mark testifying to Senate a few years back when Meta was under lots of congressional scrutiny and the CTO of Facebook was calling PyTorch superusers to get their feedback while it was happening.

• There is commitment to pursuing a future with dedicated focus. Mark explictly says he wants to “keep the main thing, the main thing”. He adds “the scarcest thing in a company of this scale is focus not cash”. And he wants to focus on AI things. Makes sense to me.

• Founders who build and are technical are the best. For example, Mark built an AI assistant to help him around the house in 2016.

• All good AI requires lots of compute, LLMs or not. And computing resources cost lots of money. That spending would likely show up as “AI/Metaverse” spending on a balance sheet. Meta will acquire 35,000 H100 GPUs which will cost more than 10B dollars. They will probably buy more. And, I remember seeing this somewhere, but Meta has the most compute of any of the big tech companies. Data and compute is oil.

  From transcript: “As we develop more advanced and compute-intensive recommendation models and scale capacity for our generative AI training and inference needs, we expect that having sufficient infrastructure capacity will be critical to realizing many of these opportunities. As a result, we expect that we will invest significantly more in infrastructure over the coming years.””

• With the purchasing of raw compute comes the task of building the infrastructure, cooling, and energy around it. All of which cost money. And, they are designing their own chips and deploying it in their data centers soon. Wonderful. It costs more up front, but it will be better than relying on third parties in the long term.

• He’s been here before … almost exactly 2 years ago was the most recent time. We have seen this story before. Facebook dropped 26% after dissapointing Q1 earnings in Feb 2022 as guidance from them raised concerns about growth prospects, and monthly active users fell for the first time ever. The stock fell another 60% from there till Nov 2022 for the above reasons, and concerns about their heavy investment into the metaverse. Their AR products are competitive including the Meta Quest and the sold out Ray-Ban glasses.

• Their social media products are not going anywhere. Their product has gotten better over the past few years. They used to be really bad at serving Reels that were relevant / targeted compared to TikTok and now, they are on par if not better. How did they do that? Deploying better ML models. And what are they investing their money in? ML. How can they better personalize social media? How can they unlock the creativity of their young users to post, engage, and communicate with each other? AI-powered tools. I played around with their image generation - which is available in the search bar in instagram - and it is good at creating novel images. Investments in AI will power growth in the social media business. This sentiment was echoed in the Q1 earnings call:

  From transcript: “There are several ways to build a massive business here, including scaling business messaging, introducing ads or paid content into AI interactions, and enabling people to pay to use bigger AI models and access more compute. And on top of those, AI is already helping us improve app engagement which naturally leads to seeing more ads, and improving ads directly to deliver more value. So if the technology and products evolve in the way that we hope, each of those will unlock massive amounts of value for people and business for us over time.””

Spending money on AI, and spending money on AI with the focus and ideas he has (e.g., open source, scaling the computing farm, custom chips, research), is a great use of resources. The specific applications will emerge but at minimum it will improve their existing suite of products including advertisting, social media experience, and augmented reality products. As mentioned in the earnings call, this is a multi-year effort. Investment in meta should be taken with this time scale.