Artificial Intelligence (AI), Generative AI

Image Source: Getty Images

Artificial Intelligence

• 90% of jobs: impacted by AI automation and augmentation; Source: Fast Company

• Large-Language Models (LLMs) most-cited websites for sourcing factual information: Reddit, Wikipedia, YouTube, Google, Yelp, Facebook, Amazon, Tripadvisor; Source: Visual Capitalist

• Traditional vs. Gen vs. Agentic AI; Source: McKinsey

  • Traditional AI: solves analytical tasks faster and more efficiently than humans (classify, evaluate, predict, optimize data)

  • Gen AI: create new content (audio, code, images, text, videos) and more readily use unstructured data

  • Agentic AI: plans and executes actions against defined objectives (web search, systems, other models/agents); offers a 20X increase in productivity potential

• valuation over time top companies: OpenAI ($300 billion), xAI ($200 billion), Anthropic ($170 billion); Source: Pitchbook

• ChatGPT receives 2.5 billion prompts globally each day (2025): 330 million from American users; Source: TechCrunch

• jobs safest from AI are public-facing (low automation risk due to constant human interaction), management roles (leadership, decision-making, judgment, strategy, and team coordination remain harder to automate), and skilled trades (physical/hands-on work is less vulnerable to AI due to situational demands that involve complex and variable environments); Source: Visual Capitalist

  • Emergency Medical Technicians

  • Healthcare social workers

  • Lawyers

  • Medical & Health Services Managers

  • First-line Supervisors of Construction Trades & Extraction Workers

  • HR Managers

  • General & Operations Managers

• 95% of companies and 35% of American consumers use generative AI: productivity, problem solving, learning new things, having fun; some have privacy, accuracy, and data security concerns; other uses of genAI; Source: Bain & Company

  • researching, gathering, summarizing information

  • writing assistance (checking grammar, polishing text)

  • creative writing (creating storylines)

  • up-to-date info (news, weather)

  • generating and editing images or other creative content

  • shopping recommendations and decision support

• how people use generative AI (2025): therapy and companionship, organize life, find purpose, enhance learning, generate code, generate ideas, fun and nonsense, improve code, creativity, healthy living, interview preparation, generate images, specific search, explainers, cooking guidance, troubleshoot, personalize learning, boost confidence, email writing, explain legalese, child entertainment, corporate LLM, student essays, travel itinerary, childcare help, medical advice, navigate personal disputes, generate legal document, conversations, anti-trolling; Source: Visual Capitalist

• jobs with high AI adoption: computer and mathematical, art/design/entertainment/sports/media, educational instruction and library, office and admin support; Source: Visual Capitalist

  • jobs with low AI adoption: farming/fishing/forestry, building and grounds cleaning and maintenance, transportation and material moving, healthcare support; Source: Visual Capitalist

• AI is becoming far more intelligent (2025): performance of LLMs on standardized tests: OpenAI's Chat GPT-3.5 (2022) demonstrated strong performance on high-school-level exams (scoring in the 70th percentile on the SAT math and the 87th percentile on the SAT verbal sections); however it often struggled with broader reasoning; today's models are near the intelligence level of people who hold advanced degrees (GPT-4 can easily pass the Uniform Bar Examination that it would rank in the top 10% of test takers and it can answer 90% of questions correctly on the U.S. Medical Licensing Examination); reasoning capabilities represents the next big leap forward for AI: reasoning enhances AI's capacity for complex decision making: allowing models to move beyond basic comprehension to nuanced understanding and the ability to create step-by-step plans to achieve goals; for businesses: fine-tune reasoning models and integrate them with domain-specific knowledge to deliver actionable insights with greater accuracy; Source: McKinsey

• AI models currently exceed human performance in almost every technical task (image classification, visual reasoning, medium-level reading comprehension, English language understanding, multitask language understanding, competition-level mathematics, PhD-level science questions) but underperform humans in multimodal understanding and reasoning (processing and reasoning across multiple formats and disciplines: charts, maps, tables, images, diagrams); Source: Visual Capitalist

• most-wanted skills for AI jobs (2024): python, computer science, data analysis, SQL, data science, automation, project management, Amazon Web Services, Agile methodology, scalability; Source: Visual Capitalist

  • Python (programming language): simple and widely-used in developing, testing, and deploying AI systems

• procurement functions that use AI can reduce overall costs by 15% - 45% and eliminate up to 30% of the work for employees and teams (these gains enable companies to reallocate procurement team capacity to more strategic, value-added tasks); Source: Boston Consulting Group

• Artificial general intelligence (AGI): theoretical AI system will capabilities that rival those of a human; purely theoretical; someday AGI may replicate human-like cognitive abilities: reasoning, problem solving, perception, learning, and language comprehension; Turing test: first proposed by 20th century computer scientist Alan Turing (when AI's abilities are indistinguishable from those of a human) - far away from reaching a point where AI tools can understand, communicate, and act with the same nuance and sensitivity of a human - and, critically, understand the meaning behind it; critical difference between current AI (2025) and AGI: essentially prediction machines that can predict with a high degree of accuracy, the answer to a specific prompt because they've been trained on huge amounts of data - but it's not a human level of performance in terms of creativity, logical reasoning, sensory perception, and other capabilities); by contrast AGI tools could feature cognitive and emotional abilities (empathy) indistinguishable from those of a human; Source: McKinsey

• "hallucinations": when AI-powered tools and applications generate incorrect information; occurs when a large language model (LLM) such as a generative AI chatbot or computer vision tool detects patterns or objects that do not exist or are imperceptible to humans: leading to outputs that are inaccurate or nonsensical; measuring hallucination rates is becoming increasingly critical as AI systems are deployed in high-stakes applications across fields such as medicine, law, finance; Source: Visual Capitalist

• most popular AI tools by website visits (2024): ChatGPT (60.2% of industry), Character.AI (15.8% of industry), QuillBot (4.7% of industry), Midjourney (2.1% of industry), and Hugging Face (1.3% of industry); Source: Visual Capitalist 

Generative AI

• companies dominating generative AI: OpenAI (Microsoft-backed) 56.3%, Microsoft (includes Github): 18.5%, Google (3.2%); Source: UNCTAD

• Anthropic Economic Index: AI's effects on the labor market and broader economy over time (how AI is being incorporated into real-world tasks across the modern economy): usage is concentrated in software development and technical writing tasks (1/3 of occupations see AI use in at least a quarter of their associated tasks); AI use leans more toward augmentation (AI collaborates with and enhances human capabilities) vs. automation (where AI directly performs tasks); AI use is more prevalent for tasks associated with mid-to-high wage occupations like computer programmers and data scientists and lower for both the lowest- and highest-=paid roles (likely reflects both the limits of current AI capabilities as well as practical barriers to using the technology); Source: Anthropic

• 72% of U.S. business leaders used Generative AI for work (2024): purchasing/procurement, product development/engineering, finance/accounting functions; Source: eMarketer

• work-related tasks where U.S. adults have used Generative AI for assistance (2024): communications or writing, brainstorming/idea generation/creativity, learning about a new topic, summarizing or simplifying longer documents, problem solving, analyzing complex information; Source: eMarketer

• Generative AI: category of AI algorithms that generate new outputs based on the data they have been trained on; unlike traditional AI systems that are designed to recognize patterns and make predictions, generative AI creates new content in the form of images, text, audio, and more; type of deep learning called generative adversarial networks (GANs) to create new content; a GAN consists of 2 neural networks: a generator that creates new data and a discriminator that evaluates the data; the generator and discriminator work together with the generator improving its outputs based on the feedback it receives from the discriminator until it generates content that is indistinguishable from real data; Generative AI has the potential to disrupt several industries including: advertising, art and design, entertainment; Source: Visual Capitalist 

• how you can tell if photo is AI generated: it takes 13 milliseconds to process a picture; AI tends to have difficulty creating eyes and hands that look real (people in deepfake videos rarely blink because AI are often trained with images of people with their eyes open) and struggles with the physics of light and shadows (including reflections); Source: National Geographic