This course reaches its final chapter at last. Here, we'll take a sweeping look at what's happening on the front lines right now. Going multimodal, reasoning models, AI agents, open source, robotics, and AGI — let's grasp the "current position" of a world whose scenery changes every six months, without jargon. Read the numbers and proper nouns as a snapshot as of March 2026.
A 5-minute overview of AI's "now"
Feeling the speed of AI's evolution
As of March 2026, when this chapter is being written, "the common sense of six months ago no longer applies" has become the watchword of the AI industry.
The numbers convey the sense of speed well. AI-related investment hit a record $225.8 billion (about ¥34 trillion) in 2025[1]. 77% of companies have adopted or are testing AI, and 21% of the world's population uses AI tools every day. The AI market size in 2025 is also estimated at roughly $244–391 billion.
about ¥34 trillion
Let's look back over the past year and a half through the major events, in chronological order.
From here, we'll dig into four especially important trends.
Multimodal AI — AI with senses
Multimodal AI means AI that can handle multiple formats in an integrated way — text, images, audio, video, and more. Early LLMs were "things that could only read and write text," but today's AI can see photos, hear voices, and make videos.
The 2025 breakthroughs
| Field | Service | What it can do |
|---|---|---|
| Image generation | GPT-4o (native image generation) | Accurately generates images with text in them. Right after its March 2025 release, demand was so high Altman described the GPUs as "melting" |
| Video generation | Google Veo 3 | Generates video with audio. Since its release, over 270 million videos have been generated in total |
| Long-context understanding | Gemini 2.5 Pro | Processes 1 million tokens (more than a full book) at once. Debuted at No. 1 on LMArena |
| Voice conversation | GPT-4o Advanced Voice | Natural real-time voice conversation without going through text. Can even be used like an interpreter |
Meanwhile, OpenAI's video-generation AI "Sora" was announced to shut down in 2026, due to an estimated infrastructure cost of about $1 million per day and sluggish user growth (web and app end in April, the API in September). It also throws into relief the reality that high-quality video generation still costs an enormous amount.
💡 A practical hint: Image analysis (photo → text) is available in almost every company's free tier. Try it in everyday life — reading receipts, turning handwritten notes into text, extracting numbers from charts, and so on.
The reasoning-model revolution — the arrival of AI that "thinks"
From late 2024, a new category was born in the world of AI: reasoning models.
Conventional AI had a "answer instantly when asked" style. Reasoning models are different. They take "time to think" before answering. It's the same as when a human solving a math problem doesn't just write the answer straight away, but thinks step by step while taking notes.
Why it matters
With the arrival of reasoning models, performance improved dramatically in the areas AI was bad at — math, science, and complex programming.
Achieved 92.7% on math-olympiad-level problems (AIME 2025). 99.5% when combined with Python tools.
Achieved high performance at a training cost of just about $6 million (compared with GPT-4's estimated $100 million-plus), reaching No. 1 in the US on the iOS App Store in January 2025. Nvidia's stock briefly fell 18%[2].
A unique feature letting developers freely set a "thinking budget." Implements "interleaved thinking," continuing to think even while using tools.
⚡ Point to watch: inference-time compute
The discovery that "letting it spend more time thinking yields more accurate answers" added a new axis to AI's evolution. In addition to the conventional "increase training data" and "make the model bigger," it turned out that increasing the amount of computation at inference time also improves performance.
AI agents — toward an era of "delegating"
The hottest keyword of 2025–2026 is AI agents.
Until now, AI was an advisor that "answers when you ask." AI agents are different. Tell it a goal, and it makes a plan, uses tools, and carries out the task autonomously. It's close to the feeling of "delegating" work to a secretary or assistant. If you want to know the mechanism in more detail, see "What is an AI agent," and for how multiple AIs work in coordination, "What is multi-agent" as well.
Concrete examples of AI agents
| Agent | What it can do | Characteristics |
|---|---|---|
| Claude Code | Autonomously handles code generation, execution, and debugging end to end | One of three products to reach $1 billion+ ARR as a coding AI |
| Operator | Operates a web browser to handle bookings and research on your behalf | Sets human checkpoints, but defending against prompt attacks is a challenge |
| Manus AI | Executes complex tasks asynchronously in the cloud | Appeared in 2025. Meta acquired it for about $2 billion, but in 2026 Chinese authorities blocked it, and unwinding procedures are underway |
| Devin | An AI engineer that autonomously carries out software development | $500/month. Official success rate is 13.86% — still developing |
🛠 If you want to seriously use a code-writing "partner," our sister course "AI Coding in Practice" explains everything from getting started with Claude Code and the like to mastering it. For people who want to build apps solo, we also recommend "Indie Development with AI."
MCP — the "common language" of AI agents
MCP (Model Context Protocol), developed by Anthropic as a standard for agents to connect with external tools, is spreading rapidly. Donated to the Linux Foundation in December 2025, its monthly SDK downloads reached 97 million. Major platforms — ChatGPT, Gemini, VS Code, AWS, Azure, and more — have all adopted it.
Gartner predicts that by the end of 2026, AI agents will be built into 40% of enterprise applications[1].
⚠️ The challenges of agents: Agents are convenient, but at present they have several important limitations. Complex judgment errors, security risks (such as sending information on their own), cost (repeatedly making API calls autonomously), and murky accountability. The iron rule is "delegate and verify," not "delegate and forget."
The rise of open-source AI
Commercial AI like GPT-4 and Claude isn't everything. Open-source AI — free to use and free to modify — is evolving at an incredible pace.
Major models (as of 2025)
| Model | Developer | Characteristics |
|---|---|---|
| Llama 4 Scout / Maverick | Meta | Scout: ultra-long-context support of 10 million tokens, runs on a single H100. Maverick: performance rivaling GPT-4o |
| DeepSeek V3 / R1 | DeepSeek (China) | V3 is GPT-4o-class at a training cost of about $6 million. R1 reached No. 1 in the US as a reasoning model |
| Qwen 3 | Alibaba | Apache 2.0 license. Supports 119 languages. Overtook Llama in downloads |
🖥 Want to run it on your own PC? Open-source models can run in your own environment too. For how to get started, "How to get started with local LLMs" is helpful, and for which model to choose, "Local LLM model comparison."
Why open source matters
Open-source AI has five significances.
You can inspect how the model works and assess its safety.
You can build a specialized model on your own data.
Run it on your own servers and API fees drop to zero.
You can use AI without sending data outside.
Prevents AI monopoly by a handful of big companies.
In the summer of 2025, a symbolic event also occurred: in total downloads of open-source models, Chinese models (DeepSeek + Qwen) overtook US-origin models. The geopolitical balance of AI development is shifting.
💡 A point for general users: Open-source AI is mainly for companies and developers, but its benefits reach everyone indirectly. As competition intensifies, commercial AI's prices fall and its performance rises. In fact, after DeepSeek R1's arrival, companies slashed their API prices.
The AI ahead — the outlook from 2026 onward
Finally, let's look at where AI is headed over the next few years through three lenses: robotics, AGI, and Japan's AI strategy.
Tesla Optimus targets a mass-production price of $20,000–30,000. Figure 03 is in live operation at a BMW plant. 1X NEO ships in 2026 for about $20,000 for home use. Chinese players have 140+ companies and 330+ models in development. A priority area in Japan's AI Basic Plan too (a solution to labor shortages).
Anthropic says early 2027, Google DeepMind within 3–5 years, OpenAI says "we know how to build it," while skeptics say "a fundamental breakthrough is still needed." The new scaling axis of investing compute at inference time is also drawing attention.
A budget of ¥1 trillion over five years. SoftBank + OpenAI invest about $41 billion, NTT tsuzumi 2 is a domestic LLM, and the Digital Agency's "Gennai" is a government AI platform. Distinctive for prioritizing innovation over regulation.
AI × robotics — putting "moving AI" into practice
Humanoid robots that combine the intelligence of LLMs with the body of a robot are finally starting to be active in real-world settings.
- Figure 03 — actually deployed at a BMW plant. Over $1 billion invested
- 1X NEO — the world's first home-use humanoid. Ships in 2026 for about $20,000 ($499/month)
- Tesla Optimus — targets a mass-production price of $20,000–30,000. Plans to produce tens of thousands in 2026
- Chinese players — 140+ companies, 330+ models in development
In Japan's government AI Basic Plan too, "physical AI" (robots × AI) is positioned as a priority area, a solution to labor shortages[3].
The road to AGI — experts' predictions
On when AGI (artificial general intelligence — AI with intelligence equal to or greater than a human's) will arrive, industry views are split. For the meaning of the term, see "What is AGI (artificial general intelligence)," and for the "technological singularity" where AI surpasses humans, we explain in detail in "What is the technological singularity."
| Position | Prediction |
|---|---|
| Anthropic | "Early 2027" — AI rivaling Nobel-laureate-level researchers by late 2026 to early 2027 |
| OpenAI | "We know how to build it" — optimistic while avoiding stating a specific date |
| Google DeepMind | "Within 3–5 years" — a big move up from its earlier "10 years" |
| Skeptical researchers | "A fundamental breakthrough is still needed" — 10–20 years on the current trajectory |
Even when people say "AGI is coming," it doesn't mean life will change dramatically overnight. Still, it's a fact that the range of what AI can do is expanding almost every month. The assumption that "AI probably still can't do this" may well be out of date six months from now.
Japan's AI strategy
In December 2025, the Japanese government adopted its first AI Basic Plan, setting out a policy to invest ¥1 trillion (about $7 billion) in AI and semiconductors over five years[3].
- SoftBank × OpenAI — SoftBank invests about $41 billion in OpenAI. The joint venture "SB OAI Japan" builds an AI platform for Japan
- NTT tsuzumi 2 — a domestic LLM. Runs on a single H100, world-class in Japanese-language performance
- Digital Agency's "Gennai" — an AI platform for about 180,000 government employees. Adopted by 7 companies including NTT, KDDI, and PFN
Japan takes a "prioritize innovation over regulation" stance, in contrast to the EU's strict regulatory approach.
What to do "now" in the age of AI
Actually experience it with free AI tools. One try beats a hundred hearsays.
AI is a tool. It becomes valuable combined with your expertise and creativity.
An era where common sense changes every six months. Staying curious is the greatest skill.
- AI has gone multimodal, able to handle not just text but images, audio, and video.
- With reasoning models, AI that "thinks" became real, and with AI agents, AI you can "delegate" to.
- Open-source AI is surging, and the geopolitical balance is shifting too.
- The next battlegrounds are robotics, AGI, and national strategies. Common sense changes every six months.
- What matters are three things: touch it and get used to it, combine it with your strengths, and enjoy the change.
References
- Gartner. "Worldwide AI Spending Will Total $1.5 Trillion in 2025." Gartner Newsroom, September 2025. / Fortune Business Insights. "Artificial Intelligence Market Report." 2025.
- "DeepSeek R1: Open-source reasoning model." DeepSeek API Docs, January 20, 2025. / Market impact reported by multiple financial outlets, January 27, 2025.
- "Japan adopts first AI basic plan with 1 trillion yen investment." Nikkei, December 2025. / "Japan AI Basic Plan." AI Strategy Headquarters, December 2025.
Related links:
- Hugging Face Models — a hub for open-source AI models
- LM Arena — a leaderboard for comparing AI model performance
From the fundamentals of AI to the latest trends, you've built up a systematic body of knowledge. AI evolves by the day. Building on the foundation you learned in this course, keep touching the tools while continuing to catch up on the latest information. If you want to master AI further as a partner for writing code, head to our sister course "AI Coding in Practice"; if you want to build a service solo, to "Indie Development with AI."