The first AI tool of almost every job is an LLM assistant
Before any specialized software, a latest-generation conversational assistant is already within reach in nearly every job.
An LLM (large language model) assistant such as Claude, ChatGPT, Mistral Le Chat or Gemini can read, write, summarize, translate, structure an idea or explain a topic. This versatility makes it a common building block: almost anyone can use it, whatever their job. We treat it as a baseline — a shared foundation — rather than a tool specific to a given job.
What a general-purpose LLM assistant can do
With no special setup or integration, a conversational assistant covers a wide range of cross-cutting tasks.
Write and rephrase
Draft text, adjust the tone, correct, condense or expand it.
Summarize and synthesize
Pull the essentials from a document, a conversation or scattered notes.
Translate and adapt
Move between languages and tailor a message to its audience.
Explain and learn
Break down a concept, answer questions, support skill-building.
Structure and organize
Turn ideas into an outline, an action list or a clear table.
Explore options
Generate options, angles or hypotheses to kick off thinking.
Why we don't count it as a “vertical tool”
In our job sheets we list vertical AI tools: software designed for a specific job or task. A general-purpose LLM assistant, by contrast, isn't tied to any one job — it touches almost all of them. Counting it on every sheet would add no distinctive information.
It's a foundation, not a specialty
The assistant is available everywhere in the same way. What sets a job apart are the specialized tools layered on top.
Avoiding noise in the sheets
Listing the same baseline on every job would bury the genuinely specific tools under needless repetition.
Measuring exposure more accurately
We measure a task's exposure to AI at the task level. The baseline is already part of that reading, without naming it as one more tool.
The main assistants available today
Several vendors offer a consumer conversational assistant. We list them plainly, with no ranking or commercial recommendation.
Claude
Anthropic
Conversational assistant geared toward writing, analysis and reasoning over long documents.
ChatGPT
OpenAI
Widely used general-purpose assistant that can converse, write and handle many formats.
Mistral Le Chat
Mistral AI
Conversational assistant from a European vendor, available in several languages.
Gemini
Assistant integrated into the Google ecosystem, from conversation to office help.
This list is indicative and not exhaustive; the landscape changes fast. The point is to illustrate the availability of a baseline, not to steer a choice.
Limits and care
An LLM assistant is still a tool to use with judgment. A few points of vigilance that apply to every job:
Check the facts
An assistant can produce a wrong answer stated with confidence. Cross-check sensitive information before using it.
Protect data
Don't share personal or confidential data without a suitable framework or your organization's approval.
Keep the decision
The assistant suggests; responsibility for the result and the final decision stays human.
For our approach to confidentiality and data protection, see our privacy policy.
And your job — where does it stand with AI?
Beyond this baseline, discover your tasks' exposure to AI, task by task, with our free test.
Test my job