What Does the Future of AI Hold? OpenAI’s Founder Reveals “What Will Change in the Next 10 Years”

What Does the Future of AI Hold? OpenAI’s Founder Reveals “What Will Change in the Next 10 Years”

Sam Altman, founder of OpenAI — the global leader in artificial intelligence — recently shared fascinating insights on the company’s first official podcast that have captured worldwide attention. The future vision painted by the company behind ChatGPT far exceeds our wildest imagination.

## Using AI for Parenting? A New Dad’s Real-Life Experience

Altman, who recently became a father, made a striking comment: “I can’t imagine parenting without ChatGPT.”

“Especially in those first few weeks, I was constantly asking ChatGPT questions about the baby’s behavior. Now I mostly ask about developmental milestones,” he shared.

**ChatGPT** is a conversational AI developed by OpenAI that can engage in natural, human-like conversations. When you ask it questions, it responds with detailed, expert-level answers.

Altman also shared intriguing predictions about how his children will use AI in the future:

“My kids probably won’t be smarter than AI, but they’ll be far more capable than we were growing up, and they’ll become incredibly skilled at using AI. They’ll probably look back at our current era as ‘very prehistoric.’”

However, he acknowledges potential challenges, including the possibility of children developing unhealthy relationships with AI and the need for society to establish new guidelines. Despite these concerns, he remains optimistic that AI’s benefits will far outweigh the drawbacks.

## What is “AGI”? Five-Year-Old Predictions Are Already Outdated

**AGI (Artificial General Intelligence)** refers to AI that can perform intellectual tasks across any domain, just like humans. Unlike traditional AI that specializes in specific tasks (like chess or image recognition), AGI can solve a wide range of problems like humans do.

According to Altman, current AI models have already far surpassed what experts considered “AGI” just five years ago.

“AI models are already smart enough, and they’ll continue to evolve. As a result, more people will consider us to have reached AGI each year.”

So what does he consider “**superintelligence**”?

“A system that can autonomously discover new science, or dramatically enhance people’s capabilities to enable new scientific discoveries.”

Why does he emphasize scientific discovery? Altman explains: “Scientific progress is the most important factor in improving people’s lives.” AI’s ability to enable more discoveries could lead to breakthrough advances in medicine, energy, environmental issues, and many other fields.

## When Will GPT-5 Arrive?

Regarding the highly anticipated GPT-5 release timeline, Altman predicts “probably around this summer,” though he notes the exact timing remains uncertain.

**GPT-5** is the next-generation version of **GPT-4**, which currently powers ChatGPT. GPT stands for “Generative Pre-trained Transformer” — an AI model that learns from vast amounts of text data to generate human-like text.

Interestingly, OpenAI is struggling with product naming conventions.

“Previously, we had a clear cycle of training new models and releasing them. But now models are more complex and can be continuously improved, so we’re unsure what to call ‘GPT-5.’”

Altman ultimately wants to return to simple naming like “GPT-5, GPT-6” to eliminate the need for users to navigate complex choices.

## The Revolutionary Impact of AI “Memory” Features

Altman describes ChatGPT’s “memory” function as one of his favorite new features.

**Memory functionality** allows AI to remember past conversations and retain user preferences and contexts. For example, you can simply say “create the usual document,” and it will remember the format and content you discussed previously to create the appropriate document.

“It’s incredible progress that AI can understand the context of users’ lives and accurately grasp what they want with minimal words.”

However, this convenient feature raises significant privacy concerns.

Recently, in a legal dispute with The New York Times, OpenAI was asked to retain user conversation records beyond 30 days. Altman strongly objected to this request.

“This is excessive. User privacy must be a core principle when using AI. Since ChatGPT conversations can be very personal, society needs frameworks that take privacy very, very seriously.”

## Will There Be Ads? OpenAI’s Revenue Model

While many internet services generate revenue through advertising, what will OpenAI do?

Altman doesn’t completely rule out advertising but shows extreme caution.

“ChatGPT has very high user trust, so introducing ads would require extreme care. Especially changing AI responses based on who pays more would completely destroy that trust.”

Unlike companies like Google, OpenAI prefers the simple model of “building good services and having people pay for them.”

## The $500 Billion “Project Stargate”

The biggest challenge in AI advancement is the shortage of “**compute**” resources.

**Compute resources** refer to the computational power needed to train and operate AI. Creating smarter AI requires more computational capacity.

“If people knew what we could do with more compute resources, they would want much, much more,” Altman explains.

To solve this problem, OpenAI is pursuing the massive “**Project Stargate**” initiative — an unprecedented infrastructure investment totaling $500 billion (approximately $75 trillion yen).

Officially announced by President Trump, this project involves collaboration between OpenAI, Oracle, and SoftBank to build massive AI-dedicated data centers over the next four years.

“The first construction site in Texas is incredibly massive, with thousands of people working on construction.”

## Energy Challenges: High Hopes for Nuclear Power

Operating such massive computational facilities requires enormous energy. Altman places great hope in “advanced nuclear power” as a future energy source.

**Advanced nuclear power** refers to next-generation technology that’s safer and more efficient than conventional nuclear power, including both nuclear fission and future nuclear fusion.

“While it’s difficult to physically move energy, we can convert energy into intelligence and send that intelligence around the world via the internet. This allows us to place training centers in locations with abundant energy sources.”

This is a uniquely innovative approach that contradicts the conventional wisdom of “building factories near consumption centers.”

## AI Transforming Science: The Era of New Discoveries

What excites Altman most is AI’s potential for scientific discovery.

“AI can analyze existing data to uncover new scientific insights and discover new applications for existing drugs.”

He considers **reasoning models** particularly important for this advancement.

**Reasoning models** are AI systems that “think” through complex problems step-by-step to arrive at better answers, similar to how humans solve difficult math problems by working through them systematically rather than jumping to conclusions.

“Users are willing to wait while AI ‘thinks’ for a while if it means getting excellent answers.”

This technology promises breakthrough discoveries in physics, chemistry, medicine, and other fields.

## Future Devices: What Comes After Smartphones?

OpenAI is also developing new hardware in collaboration with renowned designer Jony Ive (formerly of Apple).

“Current computers were designed for a world without AI. In a world with AI, what’s required of hardware and software is rapidly changing.”

Future devices are expected to become “more environmentally aware and understand more context about users’ lives,” enabling interaction through methods different from keyboards and screens.

This might include voice, gestures, or even direct thought-reading technology.

## Advice for 25-Year-Olds: What to Do Now

Altman offers specific advice for younger generations:

**1. Learn to Use AI Tools**
“We’ve rapidly shifted from an era where people said ‘learn programming’ to one where they say ‘use AI tools.’”

**2. Develop Human Skills**
Four skills are particularly important:
– **Resilience**: The ability to bounce back from difficulties
– **Adaptability**: The capacity to respond flexibly to change
– **Creativity**: The power to generate new ideas
– **Empathy**: The ability to understand what others need

“These skills are surprisingly learnable and will provide enormous returns over the coming decades.”

## Will AI Take Human Jobs? Or…

Regarding AI’s impact on employment — a concern for many — Altman offers an optimistic perspective.

“The number of people working at OpenAI will actually increase. However, each individual will accomplish far more than they could in the pre-AGI era.”

In other words, AI won’t replace humans but will function as a tool that dramatically extends human capabilities.

## Conclusion: Riding the Wave of Change

What emerges from this podcast is that AI represents more than just technological innovation — we’re at a major turning point that will transform society, the economy, and individual lives.

The key isn’t to fear this change but to actively learn, adapt, and leverage it. Starting now to become familiar with AI tools while simultaneously developing uniquely human qualities like creativity and empathy appears to be the key to success in the coming future.

The generations of our children and grandchildren will live in a world completely different from today. Why not start now with what we can do to help them thrive in that world?