11/12/2023
The hidden truths about ANNs, ML, DL, Generative AI and ChatGPT
The truth number 0. Artificial Neural Networks are Not Real AI, Not mimicking how the human brain works, or "learning" the rules, algorithms or models from finding patterns in historical data sets.
The truth number 1. ML is Not Real Artificial Intelligence, but predictive modeling, statistical classification algorithms and predictive data analytics.
The truth number 2. Deep Learning is Not Real Artificial Intelligence, but predictive modeling, statistical classification algorithms and predictive data analytics.
The truth number 3. Generative AI, combining various statistical algorithms to represent and process content as "hallucinating transformer NN models with attention mechanisms", is Not Real AI, but predictive modeling, statistical classification algorithms and predictive data analytics.
The truth number 4. LLMs, models with billions or even trillions of parameters, are Not Real AI, but predictive modeling, statistical classification algorithms and predictive data analytics.
The truth number 5. OpenAI's ChatGPT or Google's BERT are Not Real AI, but predictive modeling, statistical classification algorithms and predictive data analytics.
"Norvig and Russell book -and other authors-, saying that machine learning is equivalent to artificial intelligence is grossly misleading. ML is a contributing discipline of AI, just like search, reasoning, planning, decision making, natural language processing, vision, and robotics.
For instance, ML by itself cannot be intelligent because lacks reasoning, planning, logic, and doesn’t interact with the environment. ML detects patterns and performs predictions based on statistical analysis of data using math based algorithms. These algorithms are not intelligent per se".
The recent buzz and overhype around generative AI, LLMs and ChatGPT have been driven by the simplicity of new user interfaces for creating deepfakes or digitally forged images or videos, text or graphics, audio, or simple coding in a semi-automatic mode.
The principal difference from Eliza, now its textual tokens or images are transformed into numerical vectors using multiple encoding techniques, while encoding the biases and prejudices, racism and deception, disinformation and puffery contained in the training data, non-legally web-scraped from the internet.
Real AI is about reality, of reality and by reality, with the real system architecture:
Real AI Technology (RAIT) = the environment (the real world, the internet, virtual reality, etc.) + reality modeling and simulation and interaction engine [perception (sensors, the internet of things, robotics) + knowing/conception/classification + inference/reasoning/decision making (GOFAI) + learning (ML&DL&ANNs) + actuation (actuators, robotics)] + the environment (the real world, the internet, virtual reality, etc.)
To our surprise, OECD Council on 8 November 2023 adopted the new definition of AI to be incorporated in the EU’s new AI rulebook, which is in line with our real AI approach:
“An AI system is a machine-based system that, for explicit or implicit objectives, infers, from the input it receives, how to generate outputs such as predictions, content, recommendations, or decisions that [can] influence physical or virtual environments. Different AI systems vary in their levels of autonomy and adaptiveness after deployment,” reads the new definition.
https://www.linkedin.com/pulse/without-universal-ai-literacy-we-all-fail-learning-sota-abdoullaev-gu2me/
