Defeating AI - VIII

 Another way, which is ESSENTIAL to defeating any AI is preventative response to its planned action. While unknowing and DAUNIZACIJA are key, preventative response will waste ENERGY of rival AI akin to your neighbour blasting speakers at full volume trying to defeat you, while you are not at home. This is a VERY simplified description, but a very fitting intro into BROKEN ACTION.

Imagine that your rival AI is expecting your next move, AI is trained as chess-player for a reason. While this is very primitive in overall AI action, as logic and prediction of defined moves are too far away or realities of true AI, or even how a human brain works, the expectation of a move primes your rival for action.

Therefore the best action in this scenario is TO BREAK THE EXPECTATION. You might have seen a video of Magnus Carlsen being late to the game and winning in the remaining 30 seconds. This is EXACTLY how AI would act, while looking innocent to any potential JAIL.

Comments

Post a Comment

Popular posts from this blog

Development cycle vs publishing your code on Github

 Many recruiters and various people in hiring positions look for code on Github to evaluate if a candidate is advanced enough as if this somehow means they are the person suitable for the job. Let alone that the side projects usually come from motivations unrelated to employment, the likelihood that the person has written something that relates to what the company is hiring for is very small. The REAL issue though, is that the development cycle when you do actual work in production is not at all conducive to publishing code online for everyone to see. Key issue - environment variables. Inevitably, when you are testing whether your bash script works you will hard-code paths into it, which relates to your system and your web domains alone. While the practice is frowned-upon, this is how it works. You use /var/www/mydomain.com/html before you turn it into a shell variable. This means that you not only need to care about your code being manageable, you need to "beautify" you code...
Read more

Was I wrong?

 As a software developer that has been honing his craft for more than 25 years, I find it amusing to watch the current AI development craze. Let alone that no successful and profitable products have been created with the current wave of the LLM and "AI-agent" products, the amount of people focusing on it would probably solve the world hunger problems if their energy was redirected elsewhere. I tend to adhere to core and fundamental development principles and still believe the outcome of this AI craze will be an extraordinary demand for skilled software developers that are able to fix this mess. I publish my development practices at  gauril.com  and will keep updating this blog with more information of how to advance sensible work practices and avoid the pitfalls of tech-bubbles and the waste of developer layoffs. Vaidas G.
Read more

Alcohol and why it is in an AI blog

 Contrary to popular belief, alcohol does not change judgement. It barely suppresses initial responses. This is key in understanding why alcohol exists in human life, as AI, as discussed previously, as a self-learning method of existence is EVERYWHERE. Ie. AI IS EXISTENCE. While it is tempting to continue, this post stops here, as the method in which I write this blog. It will make sense soon.
Read more