Thoughts on A Crypto Trader Bot: Programming a Survival Mechanism with AI

Over the past few weeks, I have been working on a cryptocurrency trader bot, which I have discussed through this medium post and created a few instructional youtube videos giving instructions on creating a crypto trader bot in python.

This post will discuss an idea I have been considering over the past few weeks which is of immense interest to me because of the application of a natural phenomenon applied to ‘digital value’ and it’s location at the intersection of technology and business (life), primarily concerning concepts such as ‘bots’, ‘AI’, cryptocurrency, and if it is possible to program a ‘survival mechanism’ within code.

Note: for this post I am using the definition of ‘AI’ in a general way to be: a computer code/algorithm which executes a predetermined set of rules, with the ability to modify these rules based upon past performance. For example, within the realm of a crypto trader bots, a certain algorithm is developed for trading. After a number of trades, the ‘bot’ can determine a pattern of successful trades and modify its own code for future trades.

Discussion of Cryptocurrency — Blockchain Technology

Cryptocurrencies interest me for many reasons, and the ‘get rich quick’/’Lambo’/’moon’ portion is among the least interesting. The most fascinating aspects of crypto/blockchain technology is the low-cost of various ‘social experiments’ which can be run. For example, I believe that, at some point in the future, a ‘Universal Basic Income’ experiment will be conducted via crypto. The capabilities of crypto are barely beginning to be explored.

These capabilities are being unleashed due to the decrease in cost attributable to these technologies. Whereas previously to 2009, ‘money’ was directly tied to a finite resource (grain, gold) or a sovereign government (Dollar, Yuan), now any can issue their own cryptocurrency, essentially for free, maintained on a global network of computers maintained by thousands of independent operators, imperveous to ‘hackers’ and censorship.

Artificial Intelligence, Self-Contained CryptoTrader Bot & Survival Mechanism

I started this post by talking about AI, trading bots, and programming a ‘Survival Mechanism’ and I will return to that topic now. Few would argue that Artificial Intelligence (“Machine Learning”) is not the future: AI is taking over almost every industry from healthcare to investing. Equally, the capabilities and limitations of this technology is as unknown as its benefits.

I have discussed ‘twitter bots’ through this medium site previously, including applying a twitter bot to a ‘cloud’ service such as Heroku. This idea is so fascinating to me because of it’s replication of human behavior (tweeting, sharing ideas) coupled with the durability and automation of these behaviors through ‘code’.

As I explored this idea more, my thoughts began to move towards cryptocurrencies and trading bots and how the actions of a cryptocurrency trader could be automated and adapted through code. Eventually the idea became more refined through the attempt to ‘encapsulate’ these actions within a single entity (‘bot’) and deployed through the blockchain itself, removing almost all ‘points-of-failure’ conceivably encountered by a cryptocurrency trading bot.

My ‘great big idea’ which has interested me immensely is the idea of an autonomous, self contained ‘AI’ that has its own survival mechanism, which I have envisioned as a self-contained cryptocurrency trading bot deployed on a blockchain main network. Since ‘smart contracts’ are self-contained, self-executable code whose actions require ‘gas’ (ether) and are unmodifiable once deployed, it is not a far stretch to envision a cryptotrader bot who is completely self contained, deployed with a certain starting balance, which essentially becomes its ‘life force’ and then it becomes within the cryptotrader bot’s best interest to protect and grow this initial ‘life force’ (starting balance) as it’s own survival mechanism.

If the bot’s algorithm is profitable, it’s ‘life force’/balance will grow, thus enabling more possible ‘moves’ (buy/sell orders, trades) over a longer period of time. If the algorithm is not profitable, the balance will be ultimately drawn to ‘0'. and the bot will have no options (unless more funds are added). In a very real sense, in this situation, the bot becomes a ‘living’ creature with its own ‘motivation’ for maintaining its profitability.

And of course, just as in nature, Bots can have varying qualities: high risk, low-risk, varying strategies to ensure survival. Due to the low cost of implementing a bot (essentially the starting code, and a starting balance) then tens, hundreds, thousands or even millions of these bots could be ‘deployed’ with nothing limiting their viability except their own ‘starting code’ and ability to adapt, with their own ‘motivation to succeed’, also known as a survival mechanism.

Conclusion

I am not quite sure why this idea is so appealing to me. The idea of a self-contained, motivated trader bot (virus) deployed through a censorship proof blockchain main-net who is motivated to succed with an ability to adapt is so fascinating, I believe, because of the possibilities it presents.

These ideas are still very young within my mind, and many tests will need to be explored and challenges to overcome (such as how to actually deploy a trader bot through a smart contract), however, the possibilities and implications of this idea are too fascinating to ignore.

Ultimately I believe this idea can be extended to benefit the creator/deployer of the bot in ways such as, once the bot has achieved a certain amount of ‘profit’, the bot will transfer this excess crypto to its creator, and profitable ‘bot’ code can be replicated, but as this idea is still very new, these ideas must be explored at a later date.