Artificial Intelligence |
Why in News: India is experiencing a novel trend in its electoral landscape as AI-generated content takes center stage, shaping campaigns to deliver more personalized, community-oriented, and region-specific messaging than ever before.
:max_bytes(150000):strip_icc()/terms_a_artificial-intelligence-ai_asp-FINAL-ddba8ac599f3438d8064350d2ee1ae5a.jpg)
About Artificial Intelligence
AI refers to the capacity of computers or computer-controlled robots to carry out tasks typically requiring human intelligence and judgment.
- Scope:While AI systems cannot replicate the full spectrum of human capabilities, they can excel in specific tasks.
- Characteristics & Components:
- Ideal Trait:AI ideally possesses the capability to reason and take actions aimed at accomplishing specific objectives effectively.
- Component:Machine Learning (ML) serves as a subset of AI, facilitating automated learning processes.
- Deep Learning (DL): DL techniques enable automatic learning by assimilating vast amounts of unstructured data, including text, images, and video.
Aspect | Artificial Intelligence (AI) | Machine Learning (ML) | Deep Learning (DL) |
Definition | AI focuses on creating intelligent machines that can perform tasks requiring human-like intelligence. | ML enables computers to learn from data without being explicitly programmed. | DL uses deep neural networks to automatically learn and extract features from data. |
Relationship | AI encompasses ML and DL as its subsets. | ML is a subset of AI, and DL is a subset of ML. | DL is a subset of ML and a specialized form of AI. |
Learning Approach | AI mimics human intelligence through algorithms. | ML learns from data to improve performance over time. | DL uses deep neural networks to process data and improve accuracy. |
Data Processing | AI involves complex algorithms and decision-making processes. | ML focuses on statistical methods to identify patterns in data. | DL uses deep neural networks to process large datasets and extract features. |
Examples | AI applications include virtual assistants like Siri and Google’s AI-powered predictions. | ML applications include email spam filtering and self-learning algorithms. | DL applications include image analysis, sentiment-based news aggregation, and driverless car technology. |
ChatGPT |
Why in News: Cybersecurity professionals have issued alerts regarding the potential exploitation of ChatGPT for the creation of phishing emails and the development of malicious code, a process that could be executed swiftly and on a significantly greater scale.
About ChatGPT:
- ChatGPT (Generative Pre-trained Transformer) was launched by OpenAI in November 2022, as part of the GPT-3 series of extensive language models.
- Nature:It belongs to the category of artificial intelligence designed to comprehend and produce natural language text.
- Training: Trained extensively on vast textual datasets, it employs a transformer algorithm to grasp the nuances of generating text akin to human conversation.
- Text Generation:Utilizing its AI and machine learning education, ChatGPT excels in presenting information and responding to inquiries in a conversational manner, mirroring human interaction.
Gemini AI model |
Why in News: Google has recently unveiled its newest and most potent AI model, dubbed Gemini.
About Gemini AI Model:
- Multimodal Capability:Gemini represents a new era of multimodal general AI models, enabling simultaneous understanding and processing of various formats such as text, code, audio, image, and video.
- Availability:Users worldwide can access Gemini through Bard, certain developer platforms, and the latest Google Pixel 8 Pro devices.
- Coding Proficiency:Gemini possesses the ability to comprehend, elucidate, and generate high-quality code in prominent programming languages like Python, Java, C++, and Go.
- Size Options:Available in three variants – Ultra, Pro, and Nano.
- Gemini Ultra:The largest and most advanced model tailored for highly intricate tasks. Currently accessible to a limited audience including select customers, developers, partners, and safety experts for initial experimentation and feedback.
- Gemini Pro:Optimized for scaling across a diverse range of tasks and now accessible through Bard for regular users globally.
- Gemini Nano:Designed for on-device tasks, already integrated into Pixel 8 Pro, empowering new functionalities like Summarize in the Recorder app and Smart Reply via Gboard.
Large Language Models (LLMs) |
Why in News: The capability of Generative AI models to engage in “conversations” with humans stems from a component referred to as the Large Language Model, or LLM.
:max_bytes(150000):strip_icc()/large-language-model-7563532-final-9e350e9fa02d4685887aa061af7a2de2.png)
About Large Language Models (LLMs):
LLMs, a type of AI program, excel in tasks such as text recognition and generation.
- Training Data Size:LLMs are trained on vast datasets, hence the descriptor “large.”
- Machine Learning Foundation: Built upon machine learning principles, LLMs utilize transformer models, a type of neural network.
- Understanding Human Language: LLMs, in essence, are computer programs exposed to ample examples to comprehend and interpret human language or complex data types.
- Data Sources: Many LLMs are trained on internet-derived data, often comprising immense volumes of text data.
- Data Quality Impact:The quality of training samples influences the LLM’s proficiency in learning natural language, leading developers to consider curated datasets.
- Deep Learning in Action:LLMs leverage deep learning, employing probabilistic analysis of unstructured data to discern relationships between characters, words, and sentences autonomously.
- Tuning Process: LLMs undergo further training through fine-tuning or prompt-tuning to cater to specific tasks, such as question interpretation and response generation or language translation.
- Uses of LLMs:
- Versatile Applications:LLMs are versatile and can be trained for various tasks. A prominent application is generative AI, where they respond to prompts or questions by producing text.
- Example: Notably, publicly available LLMs like ChatGPT can generate essays, poems, and other textual forms based on user inputs.
Dark Net |
Why in News: During a session on Global Trends in Terrorist Financing and Terrorism at the ‘No Money for Terror’ Ministerial Conference on Counter-Terrorism Financing in New Delhi, Union Home Minister Amit Shah emphasized the necessity for unified actions to tackle the challenges arising from terrorists’ utilization of the dark net and virtual currencies such as cryptocurrency.
About Dark Net
The Dark Net, also known as the dark web, represents a clandestine section of the internet utilized for illicit activities such as drug trafficking and the exchange of explicit content, employing the hidden pathways of the onion router (ToR) to evade law enforcement surveillance.
- Location:Positioned beneath the private deep web, the Dark Net utilizes customized software and concealed networks layered atop the internet’s infrastructure.
- Encryption Strength:Due to its end-to-end encryption, the darknet poses significant challenges for law enforcement agencies investigating criminal activities conducted within its realms.
- Operation:
- Secrecy:Operating covertly, the dark web employs specialized browsers to prevent eavesdropping and traffic analysis attacks.
- End-to-End Encryption:The robust end-to-end encryption employed by the darknet enhances its resilience against penetration attempts.
- Access Methods: Access to the darknet is restricted to specialized browsers like Tor, Freenet, I2P, and Tails, ensuring anonymous browsing experiences.
- Tor (The Onion Router):An open-source software facilitating anonymous communication, Tor routes web page requests through proxy servers, obscuring the user’s IP address.
Factor | Deep Web | Dark Web |
Definition | The deep web is the part of the internet that is not indexed by search engines, but can be accessed with a username and password. | The dark web is a subsection of the deep web that is intentionally kept hidden and can only be accessed with special tools, such as the Tor browser. |
Access | The deep web can be accessed with a normal web browser, but requires a username and password to enter. | The dark web requires purpose-built software, such as the Tor browser, to access. |
Content | The deep web contains content that is not indexed by search engines, but is otherwise safe and can be accessed with a normal web browser. | The dark web is used for hosting encrypted websites and is known for illegal transactions. |
Anonymity | The deep web is not inherently anonymous, but can be used for anonymous activities if the user takes appropriate measures. | The dark web is designed to provide anonymity, making it difficult to track users or their activities. |
Use Cases | The deep web is used for legitimate purposes that require anonymity, such as online banking or email accounts. | The dark web is used for both legal and illegal activities, including illicit product and service trading. |
Size | The deep web is estimated to make up 90-95% of the internet, while the dark web is much smaller in size. | The exact size of the dark web is difficult to determine due to its decentralized and obscure nature. |
Examples | Examples of deep web content include paywalled websites, online banking, and restricted-access social media profiles. | Examples of dark web content include hacker forums, drug marketplaces, and news organization drop boxes for whistleblowers. |