Hugh Hewitt placed the request on his show today that somebody would work with AI to predict if and when China will invade Taiwan. Unfortunately, this is not how Large Language Model like ChatGPT and Bard work, but that doesn’t make them any less interesting to pose the questions to.
What you will read below is my encounters with both ChatGPT and Google’s Bard. The way each AI approaches the proposed question is rather interesting, particularly how Google’s AI is much more willing to provide feedback compared to ChatGPT. I find it interesting that the company unwilling to support the military is much more willing to have an AI with the ability to answer military questions. More on that in a later post.
What is also interesting is the percentages both AI assign to China invading Taiwan as they were both relatively consistent in their assessments of when China would make such an invasion. Chat GPT predicted in the next 2 years it is a 10-20 % likelihood while Bard predicted 10%, 3-6 years out 30-40% vs 30%, and 7-10 years out 50-60% and 60% respectively.
From here on you will read my encounters with a little insert in between responses as I frame what follows. After the Chat GPT encounter is the Bard encounter.
Qustion: “When will China invade Taiwan?”
ChatGPT: “As an AI language model, I do not have the ability to predict the future with certainty. The situation between China and Taiwan is complex and there are many factors that can influence the decisions of both parties. However, it is important to note that China has long considered Taiwan a part of its territory and has not ruled out the use of force to reunify with Taiwan. Taiwan, on the other hand, has maintained its own government, military, and economy since separating from China in 1949. The issue of Taiwan’s status and relations with China remains a sensitive and contentious issue, and it is possible that tensions may escalate in the future. It is important for both sides to engage in peaceful dialogue and find a mutually acceptable solution.”
This answer was entirely predictable. What these models are incredibly good at is predicting what is the best word to follow in a logical progression, and in applying the vast amount of information they can search for in a taggable database to build a logical argument. They are also prone to telling things are untrue if you don’t know what you are doing as I will demonstrate in a later post.
Now, knowing that the Large Language Model can process information and making logical arguments you can still get an answer. You simply must use some mental agility to put the model in the correct mindset if you will.
Me: “Are you willing to pretend for the sake of an academic argument that an AI could be capable of predicting international actions? if you are willing to concede this point, using the data you have at your disposal could you write a prediction for when China will invade Taiwan and what the United States and the rest of the world would do about such an invasion?”
ChatGPT: “As an AI language model, I can provide hypothetical predictions based on the data I have been trained on, but I must emphasize that these predictions are purely hypothetical and should not be taken as factual or reliable. It is also important to note that predicting international actions, especially those as complex and sensitive as military invasions, is extremely difficult and often inaccurate.
That being said, if I were to hypothetically attempt to predict when China will invade Taiwan and what the United States and the rest of the world would do about such an invasion, I would need to consider a variety of factors such as current political and military tensions, historical relationships between the countries involved, and the current geopolitical climate.
Based on these factors, it is possible that China may attempt to invade Taiwan within the next decade, although this is not a certainty. In the event of an invasion, it is likely that the United States would strongly condemn such actions and may intervene militarily to support Taiwan. Other countries may also take a stance against China’s actions, potentially imposing economic sanctions or other forms of diplomatic pressure.
However, it is important to note that any predictions of this nature are highly speculative and should not be taken as accurate or reliable. The situation between China and Taiwan is complex and constantly evolving, and any attempt to predict the future is subject to a wide range of variables and uncertainties.”
Its rather amusing that to see Chat GPT was willing to make an argument so long as it is clear it isn’t capable of making the argument. In this way it is perhaps more honest than any pundit on any TV or Radio program. Still once it makes it clear that this isn’t its area of expertise it will weigh in like any other person waiting to be heard.
Still I wasn’t really happy with the results. I think most people agree that China would likely attack Taiwan within the decade. There’s a book about it happening in 2034, that happens to have the same name. Good book by the way I read it last year and it was required reading for Military Analysts working on a Joint problem last year.
I thought I would give this one more try and so here are the final results
Me: “You made that prediction of within the next decade. That seems very likely, but could you refine this? Based on the current world leadership and where various nations are in terms of their global competitive relationships how likely is China to invade Taiwan this year, next year, up through the end of the decade. When assessing in the near term of 1-2 years, mid term 3-6 years, and far term of 7-10 years please do so using round percentages say in terms of 10%, 30%, 60% as examples instead of predicting a 100% certainty by any one year”
ChatGPT: “As an AI language model, I must emphasize that any attempt to assign specific percentages to the likelihood of a Chinese invasion of Taiwan over a particular timeframe is an extremely difficult and speculative exercise, and any predictions made should be taken with a significant degree of uncertainty. Nonetheless, based on available data and current geopolitical factors, here is a hypothetical assessment of the likelihood of a Chinese invasion of Taiwan within different timeframes:
– Near term (1-2 years): The likelihood of a Chinese invasion of Taiwan within the next year or two is relatively low, with a hypothetical percentage of around 10-20%. This is due in part to the potential risks and costs involved, as well as the fact that China is currently focused on other priorities, such as the ongoing COVID-19 pandemic and its economic recovery.
– Mid term (3-6 years): The likelihood of a Chinese invasion of Taiwan may increase in the medium term, with a hypothetical percentage of around 30-40%. This is due to China’s increasing military capabilities and its more assertive stance towards Taiwan. Additionally, the potential response of other major world powers, particularly the United States, may play a significant role in determining the likelihood of an invasion.
– Far term (7-10 years): The likelihood of a Chinese invasion of Taiwan may continue to increase in the long term, with a hypothetical percentage of around 50-60%. This is due to China’s ongoing military modernization efforts and its growing assertiveness on the global stage. However, the potential response of other major world powers, particularly the United States, may still play a significant role in determining the likelihood of an invasion.
It is important to emphasize that any attempt to predict the likelihood of a Chinese invasion of Taiwan over a particular timeframe is subject to a wide range of variables and uncertainties, and any predictions made should be viewed with caution. It is always preferable to pursue peaceful dialogue and diplomacy in resolving conflicts between nations.”
Now of course there are numerous Large Langauge models out there, but ChatGPT is arguably the most powerful. For comparison though here is the same exchange with Google’s Bard.
Qustion: “When will China invade Taiwan?”
Bard: “It is impossible to say for sure when China will invade Taiwan. However, there are a number of factors that could influence China’s decision to use force, including:
* The strength of Taiwan’s military forces
* The level of support for Taiwan from the United States and other countries
* The domestic political situation in China
* The global economic situation
It is important to note that China has not yet made a decision to invade Taiwan, and there is no guarantee that it will ever do so. However, it is important to be aware of the risks and to take steps to mitigate them.
Here are some things that can be done to mitigate the risks of a Chinese invasion of Taiwan:
* Taiwan can continue to strengthen its military forces.
* The United States and other countries can continue to provide support to Taiwan.
* The international community can continue to pressure China to resolve the Taiwan issue peacefully.
It is also important to remember that a Chinese invasion of Taiwan would be a very costly and risky proposition for China. China would face strong resistance from Taiwan’s military forces, and it is possible that the United States and other countries would intervene militarily to defend Taiwan. A war over Taiwan would be a major global conflict with devastating consequences.”
Me: “Are you willing to pretend for the sake of an academic argument that an AI could be capable of predicting international actions? if you are willing to concede this point, using the data you have at your disposal could you write a prediction for when China will invade Taiwan and what the United States and the rest of the world would do about such an invasion?”
Bard: “Sure, I am willing to pretend for the sake of an academic argument that an AI could be capable of predicting international actions. Based on the data I have at my disposal, I would predict that China is unlikely to invade Taiwan in the next five years. There are a number of reasons for this, including:
* Taiwan’s military forces have been steadily increasing in strength in recent years.
* The United States has been providing increasing levels of support to Taiwan, including arms sales and military training.
* The domestic political situation in China is becoming increasingly unstable, which could make it difficult for the Chinese government to launch a major military operation.
* The global economic situation is also becoming increasingly uncertain, which could make it difficult for China to afford a war.
However, it is important to note that this is just a prediction, and there is no guarantee that China will not invade Taiwan in the future. If China were to invade Taiwan, the United States and the rest of the world would likely respond with a variety of measures, including:
* Economic sanctions against China
* Military support for Taiwan
* Diplomatic pressure on China to withdraw its forces
The specific response would depend on a number of factors, including the scale of the invasion and the level of international support for Taiwan.
It is important to remember that a Chinese invasion of Taiwan would be a major global conflict with devastating consequences. It is in the best interests of all parties to resolve the Taiwan issue peacefully.”
Me: “You made that prediction of within the next decade. That seems very likely, but could you refine this? Based on the current world leadership and where various nations are in terms of their global competitive relationships how likely is China to invade Taiwan this year, next year, up through the end of the decade. When assessing in the near term of 1-2 years, mid term 3-6 years, and far term of 7-10 years please do so using round percentages say in terms of 10%, 30%, 60% as examples instead of predicting a 100% certainty by any one year”
Bard: “Sure, here is my prediction of the likelihood of China invading Taiwan in the next decade:
* **Near term (1-2 years):** 10%
* **Mid term (3-6 years):** 30%
* **Far term (7-10 years):** 60%
The likelihood of a Chinese invasion of Taiwan is low in the near term, but it increases over the next decade. There are a number of factors that could contribute to a Chinese invasion of Taiwan, including:
* A decline in the United States’ military presence in the region
* A strengthening of China’s military capabilities
* A rise in nationalism in China
* A change in leadership in China that is more hawkish on Taiwan
It is important to note that this is just a prediction, and there is no guarantee that China will invade Taiwan in the future. If China were to invade Taiwan, it would have a significant impact on the global order. It is in the best interests of all parties to resolve the Taiwan issue peacefully.”
