BEFORE YOU PRESS PLAY
What is discussed
AskMonty AI editorial noteSource consulted:
Montemagno begins with a common obstacle: AI agents promise to carry out practical work, but technical setup can discourage people who do not program. This Hostinger-sponsored episode presents a service he considers especially accessible. The example is a LinkedIn assistant that helps research topics, draft content and publish a post. Its appeal is the ability to begin with a conversation and capabilities that are already configured, reducing the need to choose models or connect technical infrastructure manually. Montemagno frames the exercise as something professionals in many different fields could adapt to their own work. The interface resembles a familiar conversational assistant but includes predefined capabilities for activities such as writing a blog post, developing a value proposition and reviewing pricing. Rather than expecting a perfect instruction, it asks about the topic, audience, tone and desired length. Montemagno highlights this reversal: answering relevant questions is often easier than specifying every requirement unaided. Files, images and examples can supply additional context. Application connections then allow the assistant to work with authorised services. LinkedIn is the main example, while he also explores the process for connecting Gmail. The demonstrated service offers different response modes and charges through credits. The simplified choice between faster and more advanced responses is intended to spare beginners repeated decisions about models and technical settings for ordinary tasks. For the practical exercise, Montemagno requests a post about working for yourself. He provides three themes: discipline makes freedom possible, nobody automatically arrives to solve an entrepreneur's problems, and indecision can be more costly than mistakes. He asks the assistant to research his style and develop those ideas in an informal, humorous voice. An existing post becomes an additional reference. The initial draft has linguistic problems and does not sound sufficiently like him. He requests better grammar, appropriate vocabulary and a less motivational tone, then asks for more irony. The demonstration therefore includes substantial iteration before producing an acceptable version. Throughout the process, Montemagno chooses the subject, reads the output, identifies weaknesses and directs revisions. He describes the AI as a coauthor: it changes how some of the writing is executed, while the human still contributes ideas, attention and judgement. He accepts the final draft as adequate for the demonstration rather than claiming it is perfect. Before publishing, he also requests a note stating that the post was written with his AI assistant. The episode's practical lesson is that an easy interface does not remove the need to review a text and decide whether it expresses what its author actually wants to say. With LinkedIn connected, a natural-language request starts publication. The operation requires permissions and more than one attempt. The agent reports shortening the text to overcome a problem, and Montemagno checks his profile to confirm the post has appeared. He then schedules a daily research task for 7 a.m., asking for five ideas aligned with his topics and positioning. The configured routine concerns research and suggestions; the demonstration does not establish that a post will be published automatically every morning. The proposed ideas remain starting points that he can reject, replace or develop according to his own editorial priorities. He extends the concept to calendar review, meeting preparation and research on potential clients. The wider message is that useful agents may become accessible without deep technical expertise, but users still decide what they want to delegate. Someone can ask for research and draft assistance while retaining the writing, or combine their own ideas with AI-generated text. For Montemagno, that freedom to choose the division of work is the most useful outcome of the demonstration.
Key points
- Lowering the technical barrier: the demonstration presents Hostinger's AI agents for people who do not want to configure APIs, models or infrastructure.
- Built-in capabilities and guided questions: templates, project files and examples help define the intended result.
- Connections and credits: the agent works through authorised applications, including LinkedIn; the demonstrated service uses credits.
- Writing still needs human editing: the draft requires corrections to grammar, tone, humour and the author's voice.
- Publishing succeeds after retries: the speaker checks the LinkedIn post and requests disclosure of the AI's contribution.
- Scheduled daily research: the configured routine should deliver five ideas at 7 a.m.; editorial choices and the level of delegation remain personal.
