AI boom/ AI spring in 2023, 2024 & more…
AI market:
2023, ~$200B (Statista)
In 2023, the service operation segment dominated the AI market.
Generative Adversarial Networks (GANs) accounted for over 74% of the generative AI market share in 2023.
Estimate AI market will get $1000B in 2027 or 2028.
Bain & Company predicts that the market for AI-related hardware and software will grow between 40% and 55% annually, reaching between $780 billion and $990 billion by 2027.
And $1,800 in 2030.
What I concern in AI right now:
– Vertical AI:
Vertical AI is a type of artificial intelligence (AI) that’s designed to address the specific needs of a particular industry:
Focuses on specific domains
Vertical AI can solve complex industry-specific problems that broad-spectrum AI can’t tackle as precisely.
Uses industry-specific data
Vertical AI uses tailored algorithms trained on private, industry-specific data to provide more relevant insights.
Builds proprietary datasets
Vertical AI companies can develop deep domain expertise and build proprietary datasets that are tailored to the unique challenges and opportunities within an industry.
Incorporates subject matter expertise
Vertical AI can incorporate subject matter expertise into the model to achieve more personalized and reliable results.
Some examples of vertical AI include:
Hippocractic.ai: An LLM solution that focuses on nondiagnostic healthcare tasks
Abridge: A generative AI platform that transcribes and summarizes medical conversations in real-time
Mayo Clinic: Uses generative AI-based enterprise search to improve information sharing
Harvey: Builds software for law firms
Hebbia: Builds software for financial institutions
The benefits of vertical AI include:
Providing more relevant and actionable insights
Improving operational performance and user satisfaction
Creating a flywheel effect that further enhances models and deepens the moat around a business
Helping less-experienced agents communicate using techniques similar to those of their higher-skilled counterparts
– AI agents:
An AI agent, or artificial intelligence agent, is a software program that uses machine learning and natural language processing to perform tasks, interact with its environment, and achieve goals:
Tasks: AI agents can perform a variety of tasks, such as answering questions, resolving issues, and multi-tasking.
Interaction: AI agents can interact with their environment and communicate in natural language.
Learning: AI agents can learn and improve their performance over time.
Autonomy: AI agents can work semi-autonomously or autonomously.
AI agents can be used in a variety of applications, including:
Customer support: AI agents can provide proactive customer support by analyzing customer behavior and preferences to predict needs and provide solutions. Sierra.ai
Personalized recommendations: AI agents can track user activity and preferences to recommend products and services.
Chatbots: AI agents can provide instant support and personalized suggestions.
AI agents can vary in complexity, from simple chatbots to advanced AI assistants.
=> Vertical AI Agents
3 levels of AI base Capability:
1. Weak AI/ Narrow AI
2. Strong AI/ AGI
3. Super AI/ ASI/ Self-aware AI
4 levels of AI base Functionality:
1. Reactive machine AI
IBM Deep Blue
Netflix Recommend
2. Limited Memory AI | Deep learning
Chatbots
Virtual Assistants
Self-driving cars
3. Theory of mind AI (emotion recognition)
Self-driving cars
4. Self-aware AI
Sophia (Hanson robotics)
7 stages of AI:
1. Stage One: Rule-Based AI Systems (Expert system)
2. Stage Two: Context Awareness and Retention Systems
3. Stage Three: Domain-Specific Mastery Systems (Narrow/ Weak AI)
4. Stage Four: Thinking and Reasoning AI Systems (Strong AI)/ Agents
5. Stage Five: Artificial General Intelligence (AGI)/ Innovators/ New inventions
6. Stage Six: Artificial Super Intelligence (ASI)/ Organizations/ Self-aware AI
7. Stage Seven: The AI Singularity
Growth Channels for AI startups:
Discord to get consumers (B2C)
ProductHunt, LinkedIn to find B2B
Tech Crunch, Axios, Mashable, Venture Beat,…
Y Combinator
1. Pioneers in AI 3.0 | Generative AI:
Chatbot: OpenAI ChatGPT, Google Gemini, Anthropic Claude, Microsoft Copilot, Meta AI, Poe, Huggingchat, Pi,
AI assistants: Apple Siri, Google Assistant, Amazon Alexa, Adept by Amazon, Inflection Pi, Milo
AI Chinese startups: 01.ai, Baichuan AI, Zhipu AI, Moonshot AI
Autonomous Vehicles: Tesla, Google Waymo, Aurora innovation, Nuro, GM Cruise, Amazon Zoox, Baidu Apollo, Intel Mobileye, Waabi, Waze
Data analytics: Palantir, Databricks, Lightup
Unique algorithm: Bytedance (Tiktok)
IoT: Samsara, C3.ai,
Precision medicine: Tempus AI, Insitro
AI biotech: Deepmind, Cradle, Inceptive
Robotic Processing Automation (RPA): UiPath, Automation Anywhere, Zapier, Forge, Axiom.ai, Adept, Google Duet AI
Humanoid robots: Figure AI, Tesla Optimus, Boston Dynamics, Nvidia Groot , 1x, Agility robotics, Unitree robotics
Translation: Grammarly, Google Translate, DeepL
Design: Canva, Figma, Galileo AI, Leonardo.ai by Canva
Robotics & Drones | Defense: DJI, Anduril, Vannevar labs, Quantum systems, Shield AI
Cybersecurity: Wiz, Hiddenlayer
AI enterprise | productivity: Notion, Asana
AI coding tools/ code generator: Codeium, Cognition Devin, Replit, Tabnine, Amazon CodeWhisperer, Magic, Phind, Warp, Github Copilot, Cursor, Cody
Voice generator/ text-to-speech: Elevenlabs, Resemble.ai, Replica, Coqui, Descript
Image generator/ text-to-image: Midjourney, Stability DreamStudio, OpenAI Dall-E 2, Nightcafe, Photoroom, Adobe Firefly, Ideogram, Lexica
Video generator/ text-to-video: Synthesia, Runway, Pika labs, OpenAI Sora, VEED.io, HeyGen, Twelve labs, Descript, CapCut
AI customer support: AssemblyAI, PolyAI, Cresta, Sierra, Zendesk, Slang.ai, Maven AGI
AI sales: Gong, Clari, Salesforce, HubSpot, Lavender
AI photo editor: Photoroom, PicsArt, Adobe, Imagen
Music generator: Suno, Musico, Splash, Boomy, Spotify, TuneFlow, Sonify
AI chatbot for recruiting: Moonhub
Writing | text generator: Writer, Anyword, Wordtune,
AI marketing: Jasper, Copy.ai, Writesonic
Ad collateral: Rocketium, Pencil
Text & data summerizations: Tome, Scholarcy, Summari
Insights & sum: Hume, Xembly.ai,
AI slide generator: Tome
AI search: Perplexity, You.com, Glean
AI healthcare: Atomic AI, Generate Biomedicines, Turbine, OpenEvidence, Gesund.ai, Ambience, DeepScribe, Hippocratic AI, Abridge, Latent, Athelas
AI gaming: Rosebud.ai, Luma.ai, Scenario, Astroblox, Epic, Roblox, Unity
AI education: Duolingo, Elsa Speak, Ello, Atypical AI, Studdy
AI media & entertainment: Character.ai, Descript, Tiktok, Chai,
AI legal: Harvey, Ironclad, EvenUp, casetext,
AI fintech: EvolutionIQ, Resistant.ai, Hebbia, Cleo
Avatar generator: Lensa AI, Colossyan, Elai, Remini
Logo generator:
Digital twins: Blackshark.ai, Tavus
AI agents | Agentic: SuperAGI, BabyAI, Open Interpreter, Agentsforce
What do Apple, Amazon, Intel do in the AI age? Apple Intelligence, Amazon Rufus, Intel Gaudi 3
2. The Generative AI infrastructure stack:
Generative AI (Foundation models & APIs): OpenAI GPT-4, Google Deepmind PaLM 2, Anthropic Claude 3, Mistral Large 2, Alibaba Qwen, Meta Llama 3, Cohere, AI21 labs, Reka AI
GPU supply: AWS, Azure, Google Cloud, Together.ai, Lambda, CoreWeave,
New architectures: Sanaka.ai, Together.ai
AI chip: Nvidia, AMD, Groq, Cerebras, ARM, TensTorrent, Xanadu, Lightmatter, Rebellions
Accelerated computing: Voltron Data
Data labelling: Scale AI, Snorkel AI, Surge
ML development & deployment | ML Ops: Databricks Mosaic, Replicate
PaaS: Replicate, Banana, Lepton AI, Modal, Baseten
Code: Codex, CodeGen
Synthetic data: Mostly AI, Hazy, Gretel
App frameworks: LangChain, Fixie
Model training & fine tuning: Fireworks.ai, Anyscale, Hugging face
LLM Ops: LangSmith, Promptlayer,
Open AI library: Hugging face