During my opening remarks for OODAcon this year, I noted several moments where the advancement of technology has taken me by surprise including the DARPA cyber grand challenge finale at Def Con and the images I was able to create with GPT. During our happy hour, former Congressman Will Hurd, who sits on the board of OpenAI, remarked that upcoming releases would represent a new opportunity for technology surprise.  Bob Gourley wrote about this inflection point last week as well.

Over the weekend, the newly upgraded and released ChatGPT felt like one of those moments. We will continue to evaluate these technologies and put them into context for our OODA Network, but here are a few fun experiments I conducted over the weekend that provide some insight into why this technology is so disruptive.

Summarizing complex content

I continuously threw content at ChatGPT and asked it for a summary and it performed almost flawlessly. Whether an analysis of my 1995 thesis on information warfare (100+ pages in length), an article on China policy from Foreign Affairs, or a news summary, I was continuously impressed with ChatGPT capabilities. In fact, for todays OODA Loop Daily Pulse, three of the summaries were written by ChatGPT instead of our analysts.

Here is the analysis it write of the Foreign Affairs article:

The article suggests that the 20th National Party Congress of the Chinese Communist Party (CCP), which was held in October, marked a significant shift in China’s political direction. At the congress, General Secretary Xi Jinping set himself up for another decade as China’s most powerful leader since Mao Zedong, and replaced his most economically literate Politburo colleagues with a group of loyalists. He also enshrined the Stalinist-Maoist concept of “struggle” as a guiding principle in the Party Charter, effectively turning the page on the economic liberalization that has characterized China’s recent history.

The article notes that at the congress, Xi was granted a third term as the CCP’s top leader, an unprecedented development that further solidifies his hold on power. However, the article suggests that perhaps even more significant was the way in which the congress served to codify a worldview that Xi has been developing over the past decade in carefully crafted official party communications. These texts offer insight into Xi’s ideology and motivations, including a deep fear of subversion, hostility towards the United States, sympathy with Russia, a desire to unify mainland China and Taiwan, and confidence in the ultimate victory of communism over the capitalist West.

The article argues that policy makers around the world should take note of Xi’s ambitions and work to constrain them through military deterrence and limits on China’s access to technology, capital, and data controlled by the United States and its allies. It also discusses recent protests in China and suggests that democracies should do more to support the Chinese people by facilitating safer means of communication.

Simple Strategies

In several venture capital books I’ve read, it has been noted that Sequoia Capital had a particular strategy around Apple, and I decided to ask ChatGPT what might be a similar strategy for AI.

“Sequoia Capital had an Aircraft Carrier Strategy around Apple, by making investments in the companies that would comprise of component parts of PC revolution. What is the component part of the next generation AI?”

It is difficult to predict exactly what the component parts of the next generation of AI will be, as the field is rapidly evolving and new technologies are being developed all the time. Some potential component parts of AI could include advanced algorithms and machine learning techniques, specialized hardware for training and running AI models, data storage and management systems, and software platforms for building and deploying AI applications. In addition, there may be a need for new forms of input and output devices, such as sensors and robotic interfaces, to enable AI systems to interact with the world. Ultimately, the specific component parts of the next generation of AI will depend on the specific applications and use cases that are developed.

Contextual Clarity

At MissionLink many years ago, Bill Crowell said cybersecurity has a thousand points of light, but no illumination. So, I asked GPT what that means.

What does it mean when someone says that the cybersecurity industry has a Thousand points of light and no illumination?

When someone says that the cybersecurity industry has “a thousand points of light and no illumination,” they are expressing frustration with the fragmented and disorganized nature of the industry. The term “a thousand points of light” refers to the many different players and stakeholders in the cybersecurity industry, including government agencies, private companies, and individual security experts. Each of these players brings their own unique perspective and expertise to the field, but the lack of coordination and collaboration among them can make it difficult to develop a comprehensive and effective approach to cybersecurity. Without a clear and unified vision, the industry is left with many individual points of light but no overall illumination or direction. This can make it challenging for organizations to navigate the complex landscape of cybersecurity and make informed decisions about their own security strategies.

Just For Fun

For a decade I’ve had a possible joke in my todo list where I would issue a US State Department Travel Advisory, but instead of for a country the advisory would be for cyberspace. I asked GPT to do it for me.

Write a travel advisory like those issue by the State Department, but instead of for a country make the advisory for cyberspace.

The State Department advises travelers to exercise caution when traveling in cyberspace. Cyber attacks and other malicious activities are common in this area, and travelers should be aware of the potential risks. It is important to protect personal and financial information, as well as to avoid clicking on links or downloading files from unknown sources. It is also advisable to use strong and unique passwords, and to regularly update security software and antivirus programs. Travelers should also be aware of local laws and regulations, and avoid engaging in activities that may be illegal in cyberspace. The State Department also advises travelers to register their trips with the Smart Traveler Enrollment Program (STEP) to receive the latest security updates and information from the U.S. embassy or consulate.

In addition to the above advice, the State Department also recommends that travelers in cyberspace take the following precautions:

• Be cautious when sharing personal information online, and avoid sharing sensitive information such as passwords, credit card numbers, and social security numbers.
• Use secure networks and avoid using public Wi-Fi, as these networks are often unsecured and can leave your information vulnerable to hackers.
• Avoid opening suspicious emails or text messages, and do not click on links or download attachments from unknown sources, as these may contain malware or other harmful software.
• Keep your security software and antivirus programs up to date, and regularly scan your devices for viruses and other threats.
• Use two-factor authentication whenever possible, as this adds an extra layer of security to your online accounts.
• Be aware of local laws and regulations in cyberspace, and avoid engaging in activities that may be illegal or prohibited.

If you experience a cyber attack or other security incident while traveling in cyberspace, immediately contact local authorities and the U.S. embassy or consulate for assistance.
By following these precautions, travelers can reduce the risks associated with traveling in cyberspace and have a safer and more enjoyable experience.

Or ChatGPT is happy to get creative with an idea.

Write a short story about a computer administrator that finds a terminal and turns out he is editing the programming that is running the simulation he is currently living within.

As the computer administrator sat at his desk, staring at the endless lines of code scrolling across his monitor, he couldn’t help but feel a sense of frustration. Despite his best efforts, he had been unable to identify the source of the network disruption that had been plaguing his company for the past week.

Just as he was about to give up, he noticed a small, inconspicuous terminal hidden behind a pile of cables in the corner of the room. Curious, he walked over and began poking around, and to his surprise, he found that he was able to access the underlying programming of the network itself.

At first, he didn’t think much of it – after all, as the administrator, he had access to most of the system’s core functions. But as he continued to explore, he realized that he wasn’t just editing the network’s programming – he was actually editing the simulation he and his colleagues were living within.

Panicked, he quickly tried to undo the changes he had made, but it was too late – the simulation had already begun to unravel. The walls around him began to shimmer and fade, and he could see the outlines of the virtual world he had been living in for the past several years.

As he watched in horror, the simulation collapsed in on itself, and he was left standing in a featureless void. All around him, the remnants of the virtual world he had known drifted aimlessly, like debris in the aftermath of a storm.

He knew that there was no going back – he had inadvertently destroyed the only world he had ever known. But in that moment, as he stood alone in the void, he couldn’t help but feel a sense of liberation. He was no longer bound by the constraints of the simulation, and for the first time in his life, he was truly free.

Disruptively Deceptive

It is easy to get enamored with the capabilities of ChatGPT as it can move so easily between complex and simple topics and contain its worldview within both realistic and fantastical bounds (talk only like a Jedi master), but I still can’t shake the feeling this is a watershed moment. It is also important to remember that the public is working with version 3.5, yet the OpenAI engineers are currently focused on version 5. We will be watching this closely.