## Step 1: Define the symbolic representation of the problem

Let's denote:
- $x_1$ as the amount of money invested in the milk industry,
- $x_2$ as the amount of money invested in the cheese industry.

The objective is to maximize earnings, which can be represented as $0.08x_1 + 0.06x_2$.

## Step 2: Translate the natural language description into semi-algebraic constraints

The constraints based on the problem description are:
1. $x_1 + x_2 \leq 30000$ (total investment not exceeding $30,000),
2. $x_1 \geq 3x_2$ (investment in milk industry must be at least three times the investment in cheese industry),
3. $x_1 \leq 25000$ (investment in milk industry not exceeding $25,000),
4. $x_1 \geq 0$ and $x_2 \geq 0$ (investments cannot be negative).

## 3: Symbolic representation in JSON format

```json
{
    'sym_variables': [('x1', 'money invested in milk industry'), ('x2', 'money invested in cheese industry')],
    'objective_function': '0.08*x1 + 0.06*x2',
    'constraints': [
        'x1 + x2 <= 30000',
        'x1 >= 3*x2',
        'x1 <= 25000',
        'x1 >= 0',
        'x2 >= 0'
    ]
}
```

## 4: Convert the problem into Gurobi code in Python

```python
import gurobi

def solve_investment_problem():
    # Create a new model
    model = gurobi.Model()

    # Define variables
    x1 = model.addVar(name="milk_investment", lb=0)
    x2 = model.addVar(name="cheese_investment", lb=0)

    # Objective function: maximize 0.08*x1 + 0.06*x2
    model.setObjective(0.08*x1 + 0.06*x2, gurobi.GRB.MAXIMIZE)

    # Constraints
    model.addConstr(x1 + x2 <= 30000, name="total_investment")
    model.addConstr(x1 >= 3*x2, name="milk_cheese_ratio")
    model.addConstr(x1 <= 25000, name="milk_investment_limit")

    # Solve the model
    model.optimize()

    # Print the solution
    if model.status == gurobi.GRB.OPTIMAL:
        print(f"Optimal investment in milk industry: ${x1.varValue:.2f}")
        print(f"Optimal investment in cheese industry: ${x2.varValue:.2f}")
        print(f"Maximal earnings: ${0.08*x1.varValue + 0.06*x2.varValue:.2f}")
    else:
        print("The model is infeasible.")

solve_investment_problem()
```