WHAT ARTIFICIAL INTELLIGENCE CANNOT DO , a grim note to the top 100 intellectuals of this planet , Part 2 - Capt Ajit Vadakayil

CHINAs CURRENT ECONOMIC GROWTH IS JUST 3.9 % -- NOT 6.2 % AS IS BEING PROJECTED..

CHINAs GROWTH WILL CONTINUE FALLING.. SOE CHICKENS HAVE COME HOME TO ROOTS..

INDIA IS CURRENTLY THIS PLANETs FASTEST GROWING ECONOMY AT 10.1 %

THE BEGGAR WESTERN NATIONS ARE FEEDING FAKE DATA TO ARTIFICIAL INTELLIGENCE SYSTEMS TO SHOW THAT THEIR ECONOMIES ARE NOT IN RECESSION AND IS SMELLING OF ROSES....

ARTIFICIAL INTELLIGENCE ALGORITHMS WERE USED BY THE JEWISH DEEP STATE TO CRASH THE FINANCIAL MARKETS IN 2008.

THE 2008 FINANCIAL CRISIS BROUGHT THE WORLD TO ITS KNEES.

DEEP STATE CAN DO IT AGAIN—IT IS A PIECE OF CAKE IF YOU KNOW HOW TO POISON DATA. IN 2008 HARDLY ANYBODY KNEW ABOUT ARTIFICIAL INTELLIGENCE METHODS ..

DATA POISONING OCCURS WHEN AN ADVERSARY MODIFIES OR MANIPULATES PART OF THE DATASET UPON WHICH A MODEL WILL BE TRAINED, VALIDATED, AND TESTED. BY ALTERING A SELECTED SUBSET OF TRAINING INPUTS, A POISONING ATTACK CAN INDUCE A TRAINED AI SYSTEM INTO CURATED MISCLASSIFICATION,

TARGETED DATA POISONING IS WHEN AN ADVERSARY INTRODUCES A ‘BACKDOOR’ INTO THE INFECTED MODEL WHEREBY THE TRAINED SYSTEM FUNCTIONS NORMALLY UNTIL IT PROCESSES MALICIOUSLY SELECTED INPUTS THAT TRIGGER ERROR OR FAILURE.

WHEN DATA IS DELIBERATELY MANIPULATED TO DECEIVE FOR BETTER OR FOR WORSE– IT POSES A GREAT RISK TO INTEGRITY OF A SYSTEM WE HAVE GROWN TO TRUST


THE RISK IS AMPLIFIED BY THE CONVERGENCE OF AI WITH OTHER TECHNOLOGIES: DATA-POISONING CAN INFECT COUNTRY-WIDE GENOMICS DATABASES, AND POTENTIALLY WEAPONIZE BIOLOGICAL RESEARCH, NUCLEAR FACILITIES, MANUFACTURING SUPPLY CHAINS, FINANCIAL TRADING STRATEGIES AND POLITICAL DISCOURSE. UNFORTUNATELY, MOST OF THESE FIELDS ARE GOVERNED IN SILOS, WITHOUT A GOOD UNDERSTANDING OF HOW NEW TECHNOLOGIES MIGHT, THROUGH CONVERGENCE, CREATE SYSTEM-WIDE RISKS AT A GLOBAL LEVEL.

DATA POISONINGCAN HAPPEN WHEN AN ATTACKER MODIFIES THE MACHINE LEARNING PROCESS BY PLACING INACCURATE DATA INTO A DATASET, MAKING THE OUTPUTS LESS ACCURATE. THE GOAL OF THIS TYPE OF ATTACK IS TO COMPROMISE THE MACHINE LEARNING PROCESS AND TO MINIMIZE THE ALGORITHM’S USEFULNESS.

ADVERSARIAL MACHINE LEARNING ATTACKS CAN BE CLASSIFIED AS EITHER MISCLASSIFICATION INPUTS OR DATA POISONING. MISCLASSIFICATION INPUTS ARE THE MORE COMMON VARIANT, WHERE ATTACKERS HIDE MALICIOUS CONTENT IN THE FILTERS OF A MACHINE LEARNING ALGORITHM. THE GOAL OF THIS ATTACK IS FOR THE SYSTEM TO MISCLASSIFY A SPECIFIC DATASET. BACKDOOR TROJAN ATTACKS CAN BE USED TO DO THIS AFTER A SYSTEMS DEPLOYMENT.

THERE ARE TWO MAIN TYPES OF ATTACKS THAT RELY ON ADVERSARIAL DATA: POISONING ATTACKS, IN WHICH THE ATTACKER PROVIDES INPUT SAMPLES THAT SHIFT THE DECISION BOUNDARY IN HIS FAVOR, AND EVASION ATTACKS, IN WHICH AN ATTACKER CAUSES THE MODEL TO MISCLASSIFY A SAMPLE.

THE MOST COMMON TYPE OF DATA POISONING ATTACK IS MODEL SKEWING, WHICH RESULTS IN THE CLASSIFIER CATEGORIZING BAD INPUTS AS GOOD ONES. THE ATTACKER POLLUTES TRAINING DATA IN SUCH A WAY THAT THE BOUNDARY BETWEEN WHAT THE CLASSIFIER CATEGORIZES AS GOOD DATA, AND WHAT THE CLASSIFIER CATEGORIZES AS BAD, SHIFTS IN HIS FAVOR

POISONING ATTACKS WORK BY FEEDING DATA POINTS INTO THESE SYSTEMS THAT SLOWLY SHIFT THE ‘CENTER OF MASS’ OVER TIME. THIS PROCESS IS OFTEN REFERRED TO AS A BOILING FROG STRATEGY. POISONED DATA POINTS INTRODUCED BY THE ATTACKER BECOME PART OF PERIODIC RETRAINING DATA, AND EVENTUALLY LEAD TO FALSE POSITIVES AND FALSE NEGATIVES, BOTH OF WHICH RENDER THE SYSTEM UNUSABLE.

CYBERCRIMINALS USE REVERSE ENGINEERING TO EXTRACT A REPLICA OF THE AI MODEL AND CARRY OUT THESE ATTACKS WHICH OFTEN GO UNNOTICED FOR A LONG PERIOD. THEREFORE, AI MODELS CAN BE SKEWED USING SOPHISTICATED TECHNOLOGIES TO PRODUCE A TARGETED RESULT.

CONTINUED TO 2--