Skip to main content

SOON COST ZERO

 

AI Costs Are Plummeting: What It Means for the Future

Introduction

The rapid advancement of artificial intelligence (AI) is not just about innovation but also about cost. In a recent interview, Sam Altman, CEO of OpenAI, highlighted a significant trend: the cost of using AI is dropping by 10 times every year. This has massive implications for businesses, governments, and individuals alike.

Understanding AI Costs

AI-related expenses can be divided into three categories:

  1. Hardware Costs: Setting up data centers, acquiring high-performance chips, and building infrastructure.
  2. Training Costs: Feeding large datasets into AI models to improve their accuracy.
  3. Usage Costs: The expense of running AI operations, measured in tokens.

Altman specifically referred to the last category—usage costs.

A Drastic Price Drop

In 2023, OpenAI charged $0.03 for 1,000 tokens. By 2024, the price fell to $5 per million tokens—a dramatic drop. This trend is expected to continue, making AI more affordable for businesses worldwide. Other AI platforms, like Google’s Gemini and France’s Mistral, have their own pricing models, but the overall trend is clear: AI is getting cheaper to use.

The Impact of Cheap AI

Lower AI costs encourage wider adoption. More companies can afford to integrate AI into their workflows, leading to increased efficiency and automation. However, creating an AI model remains expensive. In 2024, major tech firms invested heavily in AI infrastructure:

  • Meta, Microsoft, Google, and Amazon spent $223 billion on AI development.
  • This figure is expected to rise to $325 billion this year.

In contrast, China’s DeepSeek AI model had a much smaller budget of $1.6 billion, reflecting different investment strategies between global tech giants and Chinese firms.

The Cost of AI Training

Developing an AI model is a costly process. Here’s how much some major AI models have cost to train:

  • Google's Gemini: $191 million
  • OpenAI’s GPT-4: $63 million
  • OpenAI’s GPT-3: $4.6 million
  • China’s DeepSeek: $6 million

This shows that while AI usage costs are decreasing, the cost of building advanced models continues to rise.

India’s AI Strategy: Build or Buy?

For countries like India, a crucial question arises: should they develop their own AI models or rely on existing ones? While Western firms hope more businesses will adopt their models, India has expressed interest in creating a homegrown AI system to reduce dependence on foreign technology.

The Future of AI Investment

With billions pouring into AI development, the big question remains: where does it stop? Some speculate that the goal is Artificial General Intelligence (AGI)—AI that can think and reason like humans. Others wonder if we’re heading toward AI-powered humanoid robots. Unlike the space race, where the moon landing marked a clear victory, AI’s "moon landing moment" is still undefined.

As AI costs drop, its accessibility increases, potentially transforming industries at an unprecedented rate. The future of AI is not just about what it can do but how affordable it becomes for everyone.



Comments

Post a Comment

Popular posts from this blog

RAM Costs More… AI Does More… What Happens to Us?

The rise in RAM and other hardware prices is not merely a problem for consumers buying laptops or smartphones. It can also have an indirect impact on the future of engineering jobs and salaries . At first glance, hardware prices and an engineer's salary may seem unrelated. However, they are connected through one common factor: the cost of computing . Modern software companies depend heavily on servers, cloud platforms, GPUs, RAM, storage, and networking infrastructure. If these costs increase significantly, companies may become more careful about how they spend money on technology.  Will Higher Hardware Costs Reduce Engineer Salaries? Not necessarily—but they can change what companies expect from engineers . Imagine a company currently spends ₹10 crore on computing infrastructure and ₹5 crore on engineering salaries. If infrastructure costs rise substantially, the company may look for ways to improve productivity without increasing its workforce at the same rate. This could lead to...
Post a Comment
Read more

Astra Is Not AGI Yet

  Astra Is Not AGI Yet: Why Computer Use Should Not Be Confused with Artificial General Intelligence The arrival of increasingly powerful AI agents has created a new wave of excitement around Artificial General Intelligence (AGI) . Every time an AI system demonstrates a new ability—using a computer, operating a terminal, browsing the web, writing code, or completing a multi-step task—the question quickly appears: “Is this AGI?” The answer is not necessarily. Recent demonstrations of Astra's computer-use capabilities have generated significant attention because the system can interact with a computer in ways that resemble human behavior. It can navigate interfaces, use applications, execute tasks, and work with a terminal. These capabilities represent a major advancement in AI agents, but computer-use capability alone should not be considered AGI . AGI Has Not Arrived—At Least Not by a Clear, Universal Definition One of the biggest problems in the current AI landscape is that there...
Post a Comment
Read more

Engineers are better then Anthropic : HOW?

  Engineering in the Age of Powerful AI: What Students Must Do During Four Years of Engineering The rise of advanced AI systems like Anthropic’s Claude, OpenAI’s GPT models, and Google DeepMind has created anxiety among engineering students. Many are asking: “If AI can code, design, analyze, and even debug — what will engineers do?” The answer is simple but powerful: AI replaces repetitive tasks. It does not replace capable engineers. The role of engineers is evolving — from coders to AI-enhanced problem solvers, system thinkers, and decision makers. This article provides a complete four-year roadmap for engineering students to stay relevant, competitive, and future-proof in the AI era. The Reality of AI in Engineering AI can: Generate boilerplate code Suggest algorithms Debug simple errors Write documentation Automate repetitive development tasks But AI cannot: Understand unclear business requirements fully Take accountability for failures Make ethical decisions Manage production ...
Post a Comment
Read more