Advertisement

Manus Trial Experience

I would like to clarify that this is not a soft advertisement from a self-media platform. This account is purely for personal study.


Thank you, Mr. Yuan, for helping coordinate and obtain a Manus test code, and also especially thank Xiao Hong for the arrangement. The overall testing experience was really great!

Analysis of per capita revenue contribution of the top 10 tech companies

Thought process

The process documents are also clearly visible.

My result:

Annual Revenue per Employee for Top 10 Tech Companies by Market Cap (2025)

Rank
Company
Market Cap
Revenue
Employee Count
Revenue per Employee
1
NVIDIA (NVDA)
$2.697T
$130.50B
36,000
$3,625,000
2
Apple (AAPL)
$3.535T
$395.76B
150,000
$2,638,400
3
Meta Platforms (META)
$1.590T
$164.50B
74,067
$2,220,976
4
Alphabet (Google) (GOOG)
$2.110T
$350.02B
183,323
$1,909,297
5
Broadcom (AVGO)
$843.76B
$51.57B
37,000
$1,394,054
6
TSMC (TSM)
$912.05B
$91.88B
76,478
$1,201,382
7
Microsoft (MSFT)
$2.950T
$261.80B
228,000
$1,148,246
8
Tencent (TCEHY)
$627.07B
$88.82B
105,417
$842,361
9
Tesla (TSLA)
$847.39B
$97.69B
125,665
$777,384
10
Amazon (AMZN)
$2.126T
$637.96B
1,556,000
$409,999

Key Observations:

  1. NVIDIA leads with the highest revenue per employee at $3.63 million, which is nearly 9 times higher than Amazon's figure.

  2. Apple ranks second with $2.64 million revenue per employee despite having the highest market cap among all tech companies.

  3. Amazon has the lowest revenue per employee at 410K,despitegeneratingthehighesttotalrevenue(638B) among the top 10 tech companies.

  4. Meta Platforms shows strong efficiency with the third-highest revenue per employee at $2.22 million.

  5. Companies focused on semiconductor and hardware technology (NVIDIA, Apple, Broadcom, TSMC) generally show higher revenue per employee than software and service-oriented companies.

  6. Amazon's low revenue per employee reflects its labor-intensive business model with a massive workforce of over 1.5 million employees.

Data Sources:

  • Market cap data from companiesmarketcap.com as of March 2025
  • Revenue and employee count data from stockanalysis.com as of March 2025
  • TSMC revenue converted from TWD to USD using approximate exchange rate
  • All figures rounded to nearest whole number where appropriate

Travel plan for Tokyo in April

Thought process

Process document

My result:

Take a look at the table of contents first.

Tokyo Trip Guide - April 2025

Introduction

This comprehensive guide has been prepared for your Tokyo trip in April 2025. April is one of the most beautiful times to visit Tokyo, as it coincides with cherry blossom season, offering a magical atmosphere throughout the city. This guide covers everything you need to know about Tokyo's weather in April, must-visit attractions, cultural events (with special focus on cherry blossom festivals), and practical travel tips to help you make the most of your journey.

Table of Contents

  1. Weather in April
  2. Popular Attractions
  3. Cherry Blossom Viewing & Cultural Events
  4. Travel Tips & Practical Information
  5. Suggested Itinerary Ideas

The following details:

## Weather in April


April in Tokyo brings mild and pleasant temperatures, making it an ideal time to explore the city. The cherry blossoms are typically in full bloom during early April, transforming the city into a pink wonderland.


### Temperature

- **Average Temperature**: 14°C (57°F)

- **High Temperature**: 17°C (63°F)

- **Low Temperature**: 10°C (50°F)

- **Temperature Trend**: Temperatures increase throughout the month

  - Early April: Daily highs around 60°F (15.5°C), lows around 47°F (8°C)

  - Late April: Daily highs around 69°F (20.5°C), lows around 57°F (14°C)


### Precipitation

- **Chance of Rain**: Gradually decreasing from 33% to 31% throughout the month

- **Average Rainfall**: 4.4 inches (112mm) throughout the month

- **Rainfall Days**: Approximately 16 days with some precipitation


### What to Pack

- **Clothing Layers**: Essential for adapting to temperature changes throughout the day

- **Recommended Items**:

  - Light jacket or cardigan for mornings and evenings

  - Long-sleeved shirts and tops

  - Casual trousers, jeans, or cargo pants

  - Comfortable walking shoes (you'll be doing a lot of walking)

  - Light scarf (for cooler moments and temple visits)

  - Umbrella or light raincoat (for the occasional shower)


## Popular Attractions


Tokyo offers an incredible mix of traditional culture, futuristic technology, and everything in between. Here are the must-visit attractions during your April trip:


### Major Landmarks and Sights


#### Shibuya Crossing

The world's busiest intersection where massive video screens tower above as crowds of pedestrians cross in all directions at once. Visit at dusk for the most spectacular atmosphere when the neon lights create a magical ambiance. For a bird's eye view, head to the Shibuya Scramble Square tower.


#### Tokyo Skytree

Standing at 2,080 feet, this is the world's tallest tower with 360-degree observation decks offering spectacular views of the entire city. Admission costs up to ¥3,400 ($25) for combo tickets.


#### Senso-ji Temple

Tokyo's oldest and most significant Buddhist temple, built in 628, located in Asakusa. Don't miss the massive red lantern at Kaminari-mon (Thunder Gate) and the bustling shopping street (Nakamise-dōri) leading to the temple. Visit early morning or evening to avoid crowds.


#### Imperial Palace

The main residence of the Imperial Family in central Tokyo. While the palace itself is not open to the public, the Imperial East Gardens are accessible and particularly beautiful in April with spring blooms.


### Parks and Gardens (Perfect for Cherry Blossoms)


#### Shinjuku Gyoen National Garden

A 144-acre park with Japanese, French Formal, and English Landscape gardens. This is one of Tokyo's premier cherry blossom viewing spots with over 1,000 cherry trees of various varieties. Admission is 500 yen (free for children), and the park is open from 9:00 to 18:00 (until 16:30 before March 15). Note that alcoholic drinks are prohibited.


#### Ueno Park

One of Tokyo's most popular public parks and a top cherry blossom viewing location. The park offers free admission and features illumination from March 15 to April 6 (17:00 to 20:00). It's conveniently located just a few steps from Ueno Station.


#### Meguro River

About 800 cherry trees line the river, creating a stunning tunnel of blossoms. During the Nakameguro Sakura Festival, the trees are lit up in the evenings, creating magical reflections on the water.


### Neighborhoods Worth Exploring


#### Asakusa

A district with an old-fashioned feel of Tokyo, home to Senso-ji Temple and many traditional shops. Great for experiencing historic Tokyo.


#### Akihabara

Known as "Electric Town," this district is famous for electronics, anime, manga, and video game shops. A must-visit for pop culture enthusiasts.


#### Harajuku

The center of youth fashion and culture in Tokyo. Takeshita Street is lined with trendy boutiques and quirky shops.


#### Ginza

Tokyo's upscale shopping district with luxury brands and department stores. The main street becomes pedestrian-only on weekends.


## Cherry Blossom Viewing & Cultural Events


April in Tokyo is synonymous with cherry blossom season, offering numerous opportunities to experience this beautiful natural phenomenon alongside traditional cultural events.


### 2025 Cherry Blossom Forecast

- **Forecast Start Date**: March 22, 2025

- **Full Bloom Period**: Approximately March 29 - April 6, 2025

- **Source**: Japan Weather Association (as of January 29, 2025)


### Top Cherry Blossom Festivals


#### Nakameguro Cherry Blossom Festival

- **Location**: Nakameguro Canal, Meguro Ward

- **When**: Late March to early April 2025

- **Highlights**: 

  - Approximately 800 cherry trees lining a 3.8-km stretch of the canal

  - Lanterns illuminating the trees at night

  - Food stalls selling sakura-themed drinks and snacks

  - Spectacular reflections of blossoms in the canal water


#### Ueno Sakura Matsuri

- **Location**: Ueno Park

- **When**: Mid-March to early April 2025

- **Highlights**: 

  - One of Tokyo's most famous and historic cherry blossom spots

  - Avenue of cherry trees along the main path

  - Recognized in a haiku by the renowned poet Matsuo Basho


#### Sakura Festival in Chiyoda

- **Location**: Chidorigafuchi Green Way (near Imperial Palace)

- **When**: Mid-March to late April 2025

- **Highlights**: 

  - 700-meter-long cherry tree tunnel with about 230 trees

  - Illumination after sunset

  - Boat rentals available for viewing from the water


#### Bokutei Sakura-matsuri Festival

- **Location**: Sumida River promenade (from Azumabashi Bridge to Sakurabashi Bridge)

- **When**: Mid-March to mid-April 2025

- **Highlights**: 

  - Kilometer-long sakura boulevard along the Sumida River

  - Nighttime illumination of cherry blossoms

  - View of Tokyo Skytree in the background


### Traditional Festivals in April


#### Kanda Myojin Spring Festival

- **Dates**: March 29 - April 3, 2025

- **Location**: Kanda Myojin Shrine

- **Highlights**: Traditional ceremonies praying for abundant harvests and children's health and academic success


#### Asakusa Yabusame

- **Date**: April 19, 2025

- **Location**: Asakusa Shrine area

- **Admission**: 3,000 yen for Yabusame viewing (Kusajishi viewing is free)

- **Highlights**: Traditional horseback archery with archers in samurai attire shooting at targets while riding galloping horses


#### Meiji Jingu Spring Grand Festival

- **Dates**: April 27 - May 3, 2025

- **Location**: Meiji Jingu Shrine

- **Admission**: Free

- **Highlights**: Traditional Japanese performing arts including "Bugaku" (court music and dance) and "Nohgaku" (classical theater)


#### Akabane Baka Festival

- **Dates**: April 26-27, 2025

- **Location**: Area around Akabane Station

- **Highlights**: Student marching band performances, original "Baka Dance," costume parade, and Mikoshi (portable shrine) parade


### Seasonal Foods to Try in April


1. **Sakura Mochi**: Pink-colored mochi stuffed with red bean paste and wrapped in a sakura leaf

2. **Ichigo Daifuku**: Whole strawberry covered in red bean paste and wrapped in mochi

3. **Takenoko (Bamboo Shoots)**: Spring delicacy often eaten with rice

4. **Sansai (Mountain Vegetables)**: Including fuki (butterbur), udo (mountain asparagus), and warabi (bracken)

5. **Sakura-Flavored Treats**: Available at convenience stores, cafes, and Starbucks during cherry blossom season


## Travel Tips & Practical Information


### Transportation


#### IC Cards (Suica & PASMO)

- **What They Are**: Rechargeable smart cards for all trains and buses in Tokyo

- **How to Get One**: Available at most train stations for a refundable 500-yen fee

- **For Tourists**: Consider the "Welcome Suica" which doesn't require a deposit

- **How to Use**: Tap the card on the ticket gate or stand when boarding

- **Recharging**: Add money at automated ticket machines (with English guidance)


#### Train System Overview

- **Types of Trains**:

  - JR (Japan Railway): Aboveground trains, including the Yamanote Line

  - Tokyo Metro and Toei Subway: Underground subway systems

  - Private train lines: Connect to suburbs and outlying areas

- **Hours**: First trains around 5:00 AM, last trains between 23:30 and 00:30

- **Finding Routes**: Use apps like HyperDia or Jorudan


#### Train Etiquette

- Line up at designated spots on platforms

- No eating, drinking, or talking on cell phones

- Priority seats are for elderly, pregnant, or disabled passengers

- Women-only carriages are available during peak hours


### Accommodation Recommendations


#### Shinjuku

- **Best For**: Convenient location, shopping, entertainment, nightlife

- **Advantages**: Major transportation hub, abundance of restaurants and shops

- **Notable Hotels**: BELLUSTAR TOKYO, Keio Plaza Hotel, FLOWER TERRACE Higashi Shinjuku


#### Asakusa

- **Best For**: Budget travelers, traditional atmosphere

- **Advantages**: Historic district with old Tokyo feel

- **Notable Accommodations**: Cyashitsu Ryokan Asakusa, Kaminarimon Ryokan, Asakusa HomeSuites


#### Tokyo Station Area

- **Best For**: Convenience, business travelers, day trips

- **Advantages**: Central location, easy access to/from Narita Airport

- **Notable Hotels**: The Tokyo Station Hotel, Palace Hotel Tokyo


### Dining Tips


#### Dining Etiquette

- Say "itadakimasu" before starting your meal

- Hold chopsticks properly (not pointing at others or sticking upright in food)

- Slurping noodles is considered polite

- Say "gochisosama" when finished to show appreciation

- No tipping (it can be seen as offensive)


#### Restaurant Types

- **Izakaya**: Japanese pub with small dishes meant for sharing

- **Ramen Shops**: Often have ticket machines at the entrance for ordering

- **Conveyor Belt Sushi (Kaiten-zushi)**: Affordable sushi on a rotating conveyor

- **Department Store Food Halls (Depachika)**: Excellent for takeaway food and gifts

- **Convenience Stores (Konbini)**: Surprisingly good quality food at reasonable prices


### Cultural Etiquette


#### Basic Etiquette

- Remove shoes when entering homes, traditional establishments, temples, and some restaurants

- Avoid physical contact (hugging, kissing) as it might make people uncomfortable

- Keep conversations and phone calls quiet in public spaces

- No tipping (service is included in the price)

- Be punctual for appointments and tours


#### Temple and Shrine Etiquette

- Dress modestly

- Follow specific purification rituals at the entrance (washing hands and mouth)

- Be quiet and respectful

- Photography may be restricted in some areas


### Useful Japanese Phrases


#### Greetings

- おはようございます (Ohayou gozaimasu) – "Good morning" (polite)

- こんにちは (Konnichiwa) – "Hello" or "Good afternoon"

- こんばんは (Konbanwa) – "Good evening"

- ありがとうございます (Arigatou gozaimasu) – "Thank you"

- すみません (Sumimasen) – "Excuse me" or "I'm sorry"


#### Getting Around

- __はどこですか? (__ wa doko desu ka?) – "Where is __?"

- 駅はどこですか? (Eki wa doko desu ka?) – "Where is the station?"

- トイレはどこですか? (Toire wa doko desu ka?) – "Where is the bathroom?"

- いくらですか? (Ikura desu ka?) – "How much is it?"


#### Dining

- おいしい (Oishii) – "Delicious"

- メニューをください (Menu o kudasai) – "Menu, please"

- 水をください (Mizu o kudasai) – "Water, please"

- お勘定をお願いします (Okanjou o onegaishimasu) – "Check, please"


### Money and Shopping


#### Currency

- Japanese yen (¥) is the currency

- Cash is still widely used and preferred in many places

- ATMs at 7-Eleven, post offices, and some convenience stores accept foreign cards

- Credit cards are increasingly accepted at hotels, department stores, and restaurants


#### Tax-Free Shopping

- Tourists can shop tax-free at many stores

- Bring your passport when shopping

- Look for "Tax-Free" signs

- Minimum purchase amount may apply (usually around 5,000 yen)


### Connectivity


#### Wi-Fi and Internet

- Free Wi-Fi available at many cafes, restaurants, and public spaces

- Consider renting a pocket Wi-Fi device or getting a Japanese SIM card

- Major train stations and airports offer free Wi-Fi


#### Apps to Download

- Google Maps or Japan Travel by Navitime for navigation

- Google Translate with Japanese language pack downloaded

- HyperDia or Japan Transit Planner for train schedules

- Weather app for checking cherry blossom conditions


## Suggested Itinerary Ideas


### 3-Day Tokyo Highlights (Cherry Blossom Focus)


#### Day 1: Central Tokyo & Classic Sights

- **Morning**: Visit Imperial Palace East Gardens for cherry blossoms

- **Afternoon**: Explore Tokyo Station area and Ginza shopping district

- **Evening**: Chidorigafuchi Green Way for illuminated cherry blossoms


#### Day 2: Cultural Tokyo

- **Morning**: Senso-ji Temple in Asakusa

- **Afternoon**: Ueno Park for cherry blossom viewing and museums

- **Evening**: Nakameguro Canal for evening cherry blossom illuminations


#### Day 3: Modern Tokyo

- **Morning**: Shibuya Crossing and surrounding area

- **Afternoon**: Shinjuku Gyoen National Garden for cherry blossom viewing

- **Evening**: Tokyo Skytree for night views of the city


### 5-Day Tokyo Exploration


#### Day 1-3: Follow the 3-day itinerary above


#### Day 4: Tokyo's Unique Neighborhoods

- **Morning**: Harajuku and Takeshita Street

- **Afternoon**: Akihabara Electric Town

- **Evening**: Roppongi for nightlife and Tokyo Tower views


#### Day 5: Day Trip Options

- Option 1: Kamakura (Great Buddha, temples, and coastal scenery)

- Option 2: Hakone (Mount Fuji views, hot springs, and Lake Ashi)

- Option 3: Nikko (elaborate shrines and natural beauty)


## Conclusion


Tokyo in April offers a magical experience with cherry blossoms in full bloom and pleasant weather for exploring. This guide has provided you with essential information about the weather, must-visit attractions, cultural events, and practical tips to help you make the most of your trip. The city's blend of ancient traditions and cutting-edge modernity creates a unique atmosphere that's especially enchanting during cherry blossom season.


Enjoy your Tokyo adventure, and don't hesitate to adapt this guide to your personal interests and pace. The most important thing is to immerse yourself in the experience and embrace the beauty of Tokyo in spring.


---


*This guide was prepared in March 2025 based on the latest available information. Some details may change, so it's always good to verify specific opening hours, event dates, and transportation schedules closer to your trip.*

Summary of YouTube videos

Thought process

Process documentation

For example, the intermediate todo document:

The result for me:

# Summary of the Video "Deep Dive into Large Language Models like ChatGPT"


## Video Overview


This is a deep-dive video on large language model (LLM) technology created by Andrej Karpathy (former founding member of OpenAI, former Senior Director of AI at Tesla, and now founder of Eureka Labs). The video comprehensively introduces LLM technology that supports products like ChatGPT, covering the entire training process of model development, how to understand the "thinking methods" of models, and how to effectively use these models in practical applications.


## Detailed Content Summary


### Pre-training Data and Foundation


The training of large language models starts with collecting massive amounts of internet text data. These data sources are extensive, including various forms of text such as web pages, books, and code. Karpathy introduced pre-training datasets like FineWeb, which provide the model with foundational knowledge and language understanding capabilities.


Before processing these text data, tokenization is required to convert the text into a numeric sequence that the model can process. Tokenization is a key step in model training, determining how the model understands and generates text. Karpathy provided a detailed explanation of the principles and importance of tokenization, as well as the usage of tools like Tiktokenizer.


### Neural Network Architecture and Working Principles


The video delves deeply into how neural networks process tokenized data, including input-output mechanisms and internal structures. In particular, the Transformer architecture, as a core technology of modern LLMs, was explained in detail. Karpathy also mentioned the Transformer Neural Net 3D visualization tool, which helps viewers better understand this complex architecture.


The inference process is a critical part of how the model generates text. The video explains how the model predicts the next token based on existing text and how coherent text content is generated by iterating this process.


### From GPT-2 to Llama 3.1


Karpathy elaborated on the training and inference process of the GPT-2 model, mentioning the llm.c project, which aims to reproduce the GPT-2 model. Subsequently, he moved on to the more advanced Llama 3.1 model, explaining the inference mechanism of this foundation model and referring to Meta's released Llama 3 paper and Hyperbolic inference tool.


The video explained the different stages of model training, from pre-training to the transition to subsequent training. The pre-training stage mainly focuses on enabling the model to learn the basic structure and knowledge of language, while the subsequent training stage emphasizes optimizing for specific tasks and capabilities.


### Subsequent Training and Dialogue Capabilities


The subsequent training stage uses dialogue data to enhance the model's interactive capabilities. Karpathy introduced the InstructGPT paper and Supervised Fine-Tuning (SFT) techniques, which enable the model to better understand and execute user instructions.


The video also discusses the challenges faced by LLMs, such as the hallucination issue (where models generate content that appears reasonable but is actually incorrect). It also explains how models use tools and their knowledge and working memory mechanisms, all of which are key factors in enhancing model usability.


### Model's Self-awareness and Thinking Process


An interesting topic is the model's ability to recognize itself. Karpathy discusses how models understand their own capabilities and limitations, which is crucial for developing safer and more reliable AI systems.


The video emphasizes the importance of tokens to the model's thinking process, where the model reasons and "thinks" through sequences of tokens. Karpathy also delves into how tokenization affects model capabilities, explaining why models may struggle with specific tasks like spelling.


### Uneven Intelligence and Training Progression


LLMs exhibit "uneven intelligence," meaning inconsistency in performance across different tasks. While the model excels at some tasks, it performs relatively poorly on others. The causes and implications of this phenomenon are explained in detail in the video.


The progression from Supervised Fine-Tuning (SFT) to Reinforcement Learning (RL) represents an advanced stage in model training. Karpathy provides a detailed explanation of how reinforcement learning is applied in LLM training, including its principles and practical methods. He also introduces the DeepSeek-R1 model and its technical features.


### AlphaGo and Reinforcement Learning


As a classic case of reinforcement learning, AlphaGo receives special attention in the video. Karpathy discusses the significance and impact of AlphaGo, particularly highlighting the famous "Move 37," which demonstrated the creative thinking ability of AI systems.


Reinforcement Learning from Human Feedback (RLHF) is a key technology for improving model performance. The video provides a detailed explanation of how RLHF uses human feedback to improve models, enabling them to generate content that better aligns with human expectations.


### Future Prospects and Resource Recommendations


The final part of the video discusses the future development direction of LLM technology, predicting possible technological breakthroughs and application scenarios. Karpathy provides resources and methods for tracking the development of LLMs, including model ranking platforms like LM Arena.


He also introduced channels for obtaining and using LLMs, such as the AI News Newsletter and the LMStudio local inference tool. These resources are very valuable for audiences who wish to delve deeper into and apply LLM technology.


## Summary


This video comprehensively and deeply introduces the technical principles, training processes, and application methods of large language models. From basic data collection and processing, to complex neural network architectures, and then advanced training techniques, Karpathy explains all aspects of LLM technology in a clear and understandable way. The video not only covers technical details but also discusses the capabilities, limitations, and future development of the models, providing viewers with a comprehensive framework of LLM technical knowledge.


For audiences who wish to understand the technology behind AI products like ChatGPT, this video is an invaluable resource. Both technical professionals and general audiences interested in AI can gain deep insights and practical knowledge about LLM technology.

Gen-AI Application Market Analysis

Thought Process

Process Documentation

My Results:

Generative AI Landscape 2025

Overview

This visualization presents a comprehensive mapping of the generative AI ecosystem as of 2025, categorizing key players across the value chain from infrastructure to applications.

Market Context

The generative AI market has experienced explosive growth since the introduction of ChatGPT in late 2022, reaching approximately 25billionin2024andprojectedtogrowtobetween89.9 billion and $1 trillion by 2030-2034. This landscape map organizes the most influential companies shaping this rapidly evolving space.

Infrastructure Layer

At the foundation of the generative AI ecosystem lies the infrastructure that powers model development and deployment.

Hardware Providers

  • NVIDIA (92% GPU market share)
  • AMD
  • Intel
  • Google TPU

Cloud Service Providers

  • Microsoft Azure
  • AWS
  • Google Cloud
  • Oracle Cloud

Data Storage & Processing

  • Snowflake
  • Databricks
  • MongoDB
  • Pinecone
  • Weaviate

Foundation Models & Development

Full-Stack LLM Providers

  • OpenAI (GPT-4)
  • Anthropic (Claude 3)
  • Cohere
  • AI21 Labs
  • Mistral AI

Open Source Models & Frameworks

  • Meta (Llama 3)
  • Hugging Face
  • Stability AI
  • DeepSeek
  • MosaicML

Training & Fine-Tuning

  • Weights & Biases
  • Anyscale
  • Scale AI
  • Labelbox
  • Snorkel

Enterprise Applications

General Productivity

  • Microsoft Copilot
  • Google Workspace AI
  • Notion AI
  • Jasper
  • Writer

Code & Development

  • GitHub Copilot
  • Replit
  • Tabnine
  • Codeium
  • Sourcegraph

Sales & Customer Support

  • Salesforce Einstein
  • Drift
  • Intercom AI
  • Gong
  • Chorus.ai

Marketing & Content

  • Typeface
  • Regie.ai
  • Persado
  • Moveworks
  • Synthesia

Data Analysis & Business Intelligence

  • Tableau AI
  • Power BI
  • ThoughtSpot
  • Dataiku
  • H2O.ai

Industry Verticals

Healthcare

  • Tempus
  • Insilico Medicine
  • PathAI
  • Paige.AI
  • Owkin

Financial Services

  • Upstart
  • Adept AI
  • Alphasense
  • Kensho
  • Numerai

Legal

  • Harvey
  • Casetext
  • Ironclad
  • Lexion
  • Clearbrief

Media & Entertainment

  • Runway
  • Midjourney
  • Descript
  • ElevenLabs
  • Suno

Manufacturing & Design

  • Autodesk
  • Siemens AI
  • Cognite
  • Augury
  • Veo Robotics

Consumer Applications

Chatbots & Assistants

  • ChatGPT (59.7% market share)
  • Microsoft Copilot (14.4%)
  • Google Gemini (13.5%)
  • Perplexity (6.2%)
  • Claude AI (3.2%)

Image Generation

  • DALL-E
  • Midjourney
  • Stable Diffusion
  • Leonardo.AI
  • Ideogram

Video Generation

  • Runway Gen-2
  • Pika Labs
  • Synthesia
  • D-ID
  • HeyGen

Audio & Music

  • ElevenLabs
  • Suno
  • Udio
  • Resemble AI
  • Descript

Market Trends

The generative AI landscape continues to evolve rapidly, with key trends including:

  • Consolidation among AI providers, with major platform companies integrating smaller specialized AI firms
  • Increasing focus on AI governance and regulation, with the EU leading through its AI Act
  • Shift from experimentation to ROI focus as companies seek to demonstrate value from investments
  • Rise of AI agents capable of autonomous action, with 25% of enterprises expected to deploy them by 2025

Data Sources

This landscape is based on comprehensive research from multiple sources including:

  • IoT Analytics: "Generative AI Market Report 2025–2030" (January 2025)
  • FirstPageSage: "Top Generative AI Chatbots by Market Share" (March 2025)
  • Research and Markets: "Generative AI: Global Markets" (January 2025)
  • TechTarget: "The future of generative AI: 10 trends to follow in 2025" (February 2025)

Summary

Overall, the experience with Manus is pretty good, better than some products I've tried before, which may be related to the fact that they accumulated a lot of real user needs for AI when working on Monica. Because other AI Agents on the market were not practical, I also wrote some Shortcuts calling the ChatGPT API myself to handle daily tasks, but Manus feels more convenient and efficient.

I've experienced many CUA-related products before, such as:

Compared to Manus, the ones mentioned above are more like demos rather than mature and usable products.

Although some people question whether Manus might have a marketing aspect, I personally believe:

On one hand, leveraging hot topics to gain traffic is not a bad thing,Marketing itself is a reflection of a team's capabilities.
On the other hand, as long as the team can effectively take advantage of this wave of attention to iterate quickly, the growth rate will definitely be much faster than relying solely on "working in isolation."

I've seen people in various groups question technology packaging and prompt engineering. From an investor's perspective, this issue is more important because the depth of the technology directly determines the product's competitive barriers and long-term value. However, from a user’s point of view, this actually doesn't matter much. What users truly care about is whether the product can solve their problems, if it's convenient to use, and if the results are noticeable—they don’t pay too much attention to whether the underlying technology is packaged or not. Moreover, as the user base grows and commercialization capabilities improve, this can, in turn, continuously enhance the technical competitiveness.

Seeing excellence, we should aspire to it—there's certainly a lot for us to learn here!