CYBER THOUGHTS NEWSLETTER
The adoption of the idea that AI, as a tech sector, will dominate the future, has been quicker than any other in the past. It took several years for the Internet to be understood as transformative, likewise with mobile computing and, most recently, crypto. The proof is the amount of dollars that incumbents (Ex. MSFT, GOOG, AMZN, IBM, META) are willing to invest in the platform technologies, the speed at which they are transforming their own companies, and the apparent fear they are exhibiting at being leapfrogged. So because at least a certain portion of this movement is knee-jerk, doesn’t it make sense to exhibit some skepticism, or at a minimum a degree of scrutiny, when it comes to AI investing?
From an early-stage investment perspective, even if AI adoption is obvious and inevitable, investing in the space is fraught with peril. The cost to make a foundational model can be millions of dollars, and the required infrastructure to get value from it can be 3x that. The only way that anyone can generate a return on that type of expense is if they have a built-in audience, a sales channel to implement it quickly (to capture value), and a data set that creates differentiation. Only a handful of companies actually meet those requirements. We would argue that even the late-stage investors with huge funds (let alone the early-stage investors) can’t produce a venture-type return with those costs and the accompanying implied valuations.
Even those funds that do focus on the earlier stage commonly see valuations over $100M for a company that is less than a year old and has little or no revenue, and sometimes no product. While it is still possible to get a venture return with those high early prices, these types of investments are much riskier on a cost-adjusted basis. We’ve already seen some such companies, priced for perfection, be forced to raise at lower valuations in less than 18 months later because of some unforeseen risk.
And then there is the rate of change of this technology. Only a short time ago it seemed fine to power a startup with someone else’s LLM (ex. OpenAI or Anthropic). A startup could get to market quickly and concentrate on the user experience. These companies are now pejoratively known as ‘wrapper-companies’. Investors figured out that the value was in the LLM, not the experience and that the startup was now beholden to the whims of the foundation model builder. The most obvious example is in the companies that created copy editors. It turns out that it’s quite easy for the underlying model creator to release a new product, or feature, that completely obliterates your market. There were several companies that were using OpenAI for copy editing only to be made obsolete by ChatGPT.
While we love ML & AI as an investment space we remain vigilant for all of the ways that investing in a hot space can lead to suboptimal returns. We are on the lookout for companies that build their moat by understanding the customer and their problem and can use ML & AI as part of a solution rather than those that view AI as a hammer simply looking for a nail.
Below are a few of the articles that caught our attention this month. Moreover, we’ve inserted one or two sentences in italics, summarizing each article’s importance. We hope you enjoy and appreciate the material.
WHAT WE'RE READING
Here's a curated list of things we found interesting.
Cybersecurity Budgets Grow, But at a Slower Pace
The double-digit jumps of the last few years are over, but cybersecurity spending has been spared the worst of corporate cutbacks.
For a while we’ve been hearing this from our CISO advisors, that while spending isn’t at the same pace as prior years at least their headcount and budgets haven’t been reduced.
The 100 Most Influential People in AI 2023
Here’s who made the 2023 TIME100 AI list of the most influential people in artificial intelligence.
Who doesn’t love a top 100 list? Our favorite addition here is probably Grimes, but there are tons of amazing people leading the charge into AI.
Caesars Paid Ransom After Suffering Cyberattack
Caesars Entertainment paid roughly half of a $30 million ransom that hackers demanded after a cyberattack late this summer, another example of a major casino operator suffering from an attack as MGM Resorts grapples with the fallout of a recent incident.
Like a modern day Willie Sutton, hackers are going after casinos with ransomware, and it appears that they are willing to pay.
Deals that caught our eye.
Cisco to Acquire Splunk in Deal Valued at $28 Billion
Cisco Systems will acquire the cybersecurity and data analytics company for $157 a share in cash.
What we're listening to.
Super Data Science: ML & AI Podcast with Jon Krohn The latest machine learning, A.I., and data career topics from across both academia and industry are brought to you by host Dr. Jon Krohn on Super Data Science, the most listened-to podcast in the industry. In lighthearted conversation with renowned guests, Jon cuts through hype to fuel your professional impact.
We met Jon at a dinner last week and were immediately impressed with his passion for both the subject matter of AI and ML as well as his drive to educate. The podcast below show the power of ChatGPT’s embedded code interpreter and while the title says it’s for data scientists we believe that anyone can use this tool and these techniques to analyze their own data.
ChatGPT Code Interpreter: 5 Hacks for Data Scientists
On this week’s Five-Minute Friday, host Jon Krohn gives five reasons why he is so excited about ChatGPT’s Code Interpreter and walks listeners through its capabilities with a practical example.
Lytical Ventures is a New York City-based venture firm investing in Enterprise Intelligence, comprising cybersecurity, data analytics, and artificial intelligence. Lytical’s professionals have decades of experience in direct investing generally and in Corporate Intelligence specifically.