Types de requêtes que l’IA peut traiter

La taxonomie des tâches d'IA

Tâches de rédaction

import anthropic

client = anthropic.Anthropic(api_key='sk-ant-your-key-here')

response = client.messages.create(
    model='claude-opus-4-5',
    max_tokens=256,
    messages=[{
        'role': 'user',
        'content': (
            'Write a 150-word professional email declining a meeting invitation. '
            'Tone: polite and regretful. Suggest rescheduling next week.'
        )
    }]
)
print(response.content[0].text)

Tâches de résumé

import openai

client = openai.OpenAI(api_key='sk-your-key-here')

long_text = (
    'Artificial intelligence has transformed industries from healthcare '
    'to finance over the past decade. Machine learning models now diagnose '
    'diseases, detect fraud, and power recommendation engines. However, '
    'concerns about bias, privacy, and job displacement remain significant...'
)

response = client.chat.completions.create(
    model='gpt-4o',
    messages=[{
        'role': 'user',
        'content': f'Summarize the following in exactly 2 sentences:\n\n{long_text}'
    }]
)
print(response.choices[0].message.content)

Tâches de questions-réponses

import anthropic

client = anthropic.Anthropic(api_key='sk-ant-your-key-here')

document = 'Our refund policy allows returns within 30 days with receipt...'

response = client.messages.create(
    model='claude-opus-4-5',
    max_tokens=256,
    system='Answer only based on the document provided. If unsure, say so.',
    messages=[{
        'role': 'user',
        'content': f'Document:\n{document}\n\nQuestion: How many days do I have to return an item?'
    }]
)
print(response.content[0].text)

Tâches de programmation

import openai

client = openai.OpenAI(api_key='sk-your-key-here')

response = client.chat.completions.create(
    model='gpt-4o',
    messages=[
        {
            'role': 'system',
            'content': 'You are a Python expert. Include type hints and a brief docstring.'
        },
        {
            'role': 'user',
            'content': (
                'Write a function that takes a list of integers and returns '
                'the second largest unique value, or None if it does not exist.'
            )
        }
    ]
)
print(response.choices[0].message.content)

Tâches de remue-méninges

import anthropic

client = anthropic.Anthropic(api_key='sk-ant-your-key-here')

response = client.messages.create(
    model='claude-opus-4-5',
    max_tokens=512,
    messages=[{
        'role': 'user',
        'content': (
            'Generate 10 unique app name ideas for a habit-tracking app targeting busy professionals. '
            'Each name should be one or two words. Prioritize variety — avoid similar-sounding names. '
            'After each name, add a 5-word tagline.'
        )
    }]
)
print(response.content[0].text)

Tâches de traduction

import openai

client = openai.OpenAI(api_key='sk-your-key-here')

text_en = 'Please ensure the deliverables are submitted prior to the agreed deadline.'

response = client.chat.completions.create(
    model='gpt-4o',
    messages=[{
        'role': 'user',
        'content': (
            f'Translate the following into natural, idiomatic Spanish '
            f'suitable for a casual work chat message:\n\n{text_en}'
        )
    }]
)
print(response.choices[0].message.content)

Tâches de classification

import anthropic

client = anthropic.Anthropic(api_key='sk-ant-your-key-here')

reviews = [
    'The product broke after one day. Terrible quality!',
    'Shipping was fast and the item is exactly as described.',
    'It is okay, nothing special but does the job.'
]

for review in reviews:
    response = client.messages.create(
        model='claude-opus-4-5',
        max_tokens=16,
        messages=[{
            'role': 'user',
            'content': (
                f'Classify this review as exactly one of: POSITIVE, NEGATIVE, NEUTRAL.\n'
                f'Reply with only the label.\n\nReview: {review}'
            )
        }]
    )
    label = response.content[0].text.strip()
    print(f'{label}: {review[:40]}...')

Tâches d'extraction de données

import openai
import json

client = openai.OpenAI(api_key='sk-your-key-here')

email_text = (
    'Hi, I am Sarah Johnson from Acme Corp. '
    'Please send the invoice for $4,250 to billing@acmecorp.com by June 15th.'
)

response = client.chat.completions.create(
    model='gpt-4o',
    messages=[{
        'role': 'user',
        'content': (
            'Extract the following fields from the email below and output valid JSON:\n'
            'name, company, email, amount, deadline\n\nEmail:\n' + email_text
        )
    }]
)

raw = response.choices[0].message.content
data = json.loads(raw)
print(data)

Combiner plusieurs types de tâches

import anthropic

client = anthropic.Anthropic(api_key='sk-ant-your-key-here')

response = client.messages.create(
    model='claude-opus-4-5',
    max_tokens=512,
    messages=[{
        'role': 'user',
        'content': (
            'Do the following in order:\n'
            '1. Brainstorm 5 blog post titles about remote work productivity.\n'
            '2. Pick the most compelling title.\n'
            '3. Write a 3-sentence intro paragraph for that post.\n'
            'Format your output with clear numbered sections.'
        )
    }]
)
print(response.content[0].text)

Choisir le bon cadre de tâche

import openai

client = openai.OpenAI(api_key='sk-your-key-here')

pricing_text = (
    'Starter: $9/month - 1 user, 5GB storage. '
    'Pro: $29/month - 10 users, 50GB storage. '
    'Enterprise: custom pricing - unlimited users.'
)

# Extraction frame
response = client.chat.completions.create(
    model='gpt-4o',
    messages=[{
        'role': 'user',
        'content': (
            'Extract plan names and monthly prices as a JSON array. '
            'Use null for custom pricing.\n\n' + pricing_text
        )
    }]
)
print(response.choices[0].message.content)

Vérification des connaissances

Récapitulatif de la taxonomie des tâches