PortuguêsPTEnglishENEspañolES
WhatsApp

Artificial Intelligence

AI won't wait for your company.

It has advanced more in the last three years than in the six decades before. The question is no longer whether it's worth using. Now it's how to use it well, with your team ready.

How it evolved

From 1956 to now.

Decades of slow steps and then, suddenly, a rush forward. Drag or scroll to see.

  1. 1956

    The field is born

    The Dartmouth conference, organized by John McCarthy, gives artificial intelligence its name and its start as a field of study.

    Source
  2. 1997

    Deep Blue beats Kasparov

    IBM's computer beats the world chess champion 3.5 to 2.5, under official tournament time controls.

    Source
  3. 2012

    Neural networks learn to see

    AlexNet shows the power of deep neural networks to recognize images and opens the decade of deep learning.

    Source
  4. 2016

    AlphaGo wins at Go

    Google DeepMind's AI defeats Lee Sedol 4 to 1, in a game considered too hard for machines.

    Source
  5. 2017

    The Transformer arrives

    The paper that introduces the Transformer lays the groundwork for almost every language model in use today.

    Source
  6. 2022

    ChatGPT reaches the public

    On November 30, OpenAI launches ChatGPT, and generative AI becomes part of daily life for millions of people.

    Source
  7. 2023

    AI starts to see

    GPT‑4 is introduced as a model that understands images and text at the same time.

    Source
  8. 2024

    AI uses the computer

    Anthropic launches computer use: the AI operates a computer by looking at the screen, clicking and typing.

    Source
  9. 2026

    Agents do real tasks

    In tests of computer tasks, AI agents jumped from about 12% to about 66% success, close to the 72% scored by people.

    Source

The AIs on the market

The artificial intelligences moving fastest.

  • ChatGPT
  • Claude
  • Google Gemini
  • Meta AI
  • Microsoft Copilot
  • Perplexity
  • DeepSeek
  • Mistral AI
  • Midjourney
  • Runway
  • ElevenLabs
  • Hugging Face
  • ChatGPT
  • Claude
  • Google Gemini
  • Meta AI
  • Microsoft Copilot
  • Perplexity
  • DeepSeek
  • Mistral AI
  • Midjourney
  • Runway
  • ElevenLabs
  • Hugging Face
  • ChatGPT
  • Claude
  • Google Gemini
  • Meta AI
  • Microsoft Copilot
  • Perplexity
  • DeepSeek
  • Mistral AI
  • Midjourney
  • Runway
  • ElevenLabs
  • Hugging Face

The brands mentioned belong to their owners and appear here only as a reference to the platforms on the market.

The speed

Faster than the internet.

Generative AI reached more than half the population in three years, faster than the personal computer and the internet. And it got much cheaper along the way.

5x

a year: that's how fast the computing power used to train leading models has grown since 2020.

Epoch AI

What it's already changing

People who use it well get more done.

In customer service. A study of more than 5,000 support agents, published by the NBER, showed that an AI assistant increased issues resolved per hour by 14%. Among beginner agents, the gain was 34%. AI helped the least experienced work like the most experienced.

In production. In a controlled GitHub experiment, programmers with an AI assistant finished the same task about 55% faster.

In Brazil's small businesses. According to Sebrae, the main use of AI in micro and small businesses is marketing and promotion. And the biggest difficulty for MEIs (Brazilian sole proprietors) is not knowing how to apply it to their own business. That's where we come in.

The tool is public. The gain is in putting AI in the right place in your process.

Where AI fits in

AI goes where your business gets stuck.

We start with the process, choose the technology afterwards and only wrap up when your team knows how to use it.

  • Customer service. Assistants that answer, qualify and route each contact to the right person.
  • Content and creatives. More speed to produce and test, with human review and your brand's voice.
  • Data and decisions. Reports and analysis that cut manual work and make it clear what to do.
  • Training. Team training built on the company's own process, with everyday cases.

Getting the team ready

Work is already changing.

Companies that keep up are training their people now. Those who wait will compete for the same customers against faster teams.

How to start

Four steps so you don't fall behind.

You don't need to transform the whole company at once. You need to start right.

  1. 01

    Pick one bottleneck

    Start with a process that gets stuck every day: answering quotes, scheduling, producing content, putting reports together. One visible gain convinces more than a giant plan.

  2. 02

    Test small, measure the result

    Run AI on that process for a few weeks and compare time and quality with the old way. If it worked, expand. If not, adjust.

  3. 03

    Train the team on the daily work

    A tool nobody knows how to use ends up in a drawer. Training starts from the work the team already does, with real cases from the company.

  4. 04

    Keep a person in charge

    AI speeds things up, but human review protects your brand's voice and your customers' data. It works with your team, not in its place.

Sources

Further reading.

Next step

Which process could AI fix first?

A real person answers, not a bot.

Message us on WhatsAppSend us one line