Tag: ai

  • Lost in translation: How AI Can Boost Your English Communication Skills

    Lost in translation: How AI Can Boost Your English Communication Skills

    Have you ever found yourself pondering, “Are there any English errors in this sentence?” If so, you’re not alone. Navigating the intricacies of the English language can be challenging, even for native speakers. In this blog post, I will show you real examples of how I’ve been using AI as a non-native English speaker who dares to maintain humor while refining my language skills.

    Here is my main simple-yet-effective prompt:

    Are there any English errors in this sentence? What are the errors, and why?

    Let’s dive deep on a real example:

    a print of my conversation with chatGPT, the transcript:
are there any english errors on this sentence? what are the errors and why?
a life is a huge broaden THING, so this post will be big. guide yourself by the table of
contents.

Yes, there are some errors in the sentence. Here's the corrected version:
"A life is a broad and significant thing, so this post will be substantial. Guide yourself by the
table of contents."

Here are the errors and their explanations:
1. "broaden" should be replaced with "broad" because you are describing the noun "thing,"
and "broad" is the appropriate adjective in this context.
2. "HUGE" should be changed to "significant" or a similar word because "huge" is an informal
and somewhat exaggerated term. "Significant" conveys a similar meaning in a more formal
and precise manner.
3. "Guide yourself" is grammatically correct but can be improved for clarity. "Guide yourself by
the table of contents" is somewhat awkward. You can simply say "Refer to the table of
contents" for a clearer and more concise expression.
The corrected sentence maintains clarity and gramatical correctness while using more
appropriate and precise language.

    My initial phrase was:

    “a life is a huge broaden THING, so this post will be big. guide yourself by the table of contents.”

    You can see I had an overall idea of what I wanted to say, but I didn’t want to sound weird or, worse, not be comprehended. At the same time, I didn’t want to loose the sense of humor expressed on my phrase. Given that, that’s how I analyse the chatGPT answer:

    The first tip was extremely necessary, I learned something;

    The second point was exactly the thing that, if I used, I would loose the human and funny part of my sentence, so I understood AI’s tip, but I choosed not to use it;

    On the third bullet, it said that it would be awkward to use the sentence that way, that’s valuable, it means that people would understand it, but would find it a weird colocation. The sentence construction could make sense in portuguese, but not in english. Or maybe I tried to sound natural, but I didn’t succeded on it.

    That is the main thing I’m trying to spot when taking the next step in communicating in english, how to ‘sound more natural’.

    So in the end, I wrote the phrase like this:

    Life is a huge broad THING, so this post will be substantial. Navigate through it using the table of contents. (through the post, I mean)

    Simple Tweak of Words

    If you think that’s a very time-consuming way of learning, it’s because maybe it is, at least when compared with talking to a person.

    If I had the money right now, I would just hire a native English teacher; it could also be more fun. I would collect my doubts during the week and present them to my teacher in class. It would be faster to talk instead of typing. The conversation would probably go this way:

    ‘Have you ever heard ‘substantial’ used to mean ‘big?’

    ‘Yes! It didn’t occur to me; that’s a great improvement to the phrase. Thanks!’

    Then we could skip to the next.

    The AI corrections are not always that long, though. Sometimes the sentence is just almost correct:

    a print of my conversation with chatGPT, the transcript:
are there any english errors on this sentence? what are the errors and why?
you can see I had an overrall idea of what I wanted to say, but I didn't want to sound weird or,
worse, not been comprehended.

The sentence you provided contains a minor error:
"not been comprehended" should be corrected to "not be comprehended."
Here's the corrected sentence:
"You can see I had an overall idea of what I wanted to say, but I didn't want to sound weird or,
worse, not be comprehended."
The correction involves changing "not been comprehended" to "not be comprehended" to
use the correct verb form in the context of the sentence.
    I love when it says it’s just a ‘minor error’ 😮‍💨

    Also, sometimes I just want to find a better phrasing, so instead of googling ‘synonyms for …,’ I can just ask AI:

    ‘I don’t want to repeat “during” in the same sentence: ‘I would collect my doubts during the week and show them to my teacher during the class.’

    With which it would respond:

    ‘Certainly, you can rephrase the sentence to avoid repeating “during.” Here’s an alternative way to say it:

    ‘I would collect my doubts during the week and present them to my teacher in class.’

    Simple Text Correction

    Sometimes my doubts revolve around simple things like ‘should I use ‘in’ or ‘at,’ ‘on’ or ‘in,’ and this is something I can ask AI to ‘correct the phrase’ simply or, if I have the time, ‘explain why this phrase is incorrect.’

    If most of your English mistakes are in this area, you might already know the right way to say things, but even after years, it’s totally normal for non-native speakers like us to slip up, and it’s usually no big deal during casual chats.

    However, when it comes to writing, correction becomes necessary. Hiring an editor to correct texts in bulk can also be more practical and safer, since ChatGPT can sometimes change phrases (even when we BEG it not to do so) and spotting these changes can be challenging and tiresome during lengthy text corrections.


    Whether we opt to learn from AI suggestions, hire a human professional or decide to go for any other method, our primary objective when learning a new language remains clear: to be understood. Our secondary aim is to ‘sound more natural.’

    After all, even in our native language, we occasionally grapple with this aspect, depending on the context — be it work-related, interactions with customer service representatives, or in between different social circles. At times, we’re all susceptible to coming across as a bit unusual.

    Nevertheless, there are moments when our expressions hit the mark perfectly:

    🥳

    Pin this post

  • Boosting Your Obsidian Second Brain with the Smart Connections Plugin

    Boosting Your Obsidian Second Brain with the Smart Connections Plugin

    Find out how to make your note-taking system in Obsidian even more efficient with the Smart Connections plugin. This plugin uses artificial intelligence to help you find more related information among your notes in the same vault and allows you to “chat” with it.

    [ez-toc]

    How the Smart Connections Plugin Works

    Once installed, simply open any note in your Obsidian. When you open the right-side menu, the plugin will display notes that are more relevant based on identified connections.

    menu lateral direito com notas relacionadas e suas porcentagens de relação
    menu lateral direito com notas relacionadas e suas porcentagens de relação

    For example, in the image above, you can see that the plugin on the right-side menu indicates the percentage of relevance that the currently open note has with others in my vault.

    Chatting with the Plugin

    Another useful feature of “Smart Connections” is its chat capability, similar to how we chat with ChatGPT. However, it responds based on the content of your notes. For example, in the image below, you can see that I ask it to summarize a note.

    menu lateral direito com chat para consultar suas notas
    menu lateral direito com chat para consultar suas notas

    This feature is handy for handling private information that hasn’t been shared online yet. It’s also more reliable because the information comes from your own notes. As many of us have noticed, ChatGPT can occasionally give inaccurate responses, so searching within your own notes seems like a more dependable approach.

    How to Install the Smart Connections Plugin

    1. Access Obsidian settings > “Plugin Community.”
    2. Search for “Smart Connections” and click to install.
    3. To use the plugin, you’ll need an OpenAI API key to configure it. You can obtain this key on the OpenAI website and input it in the plugin settings in Obsidian.

    Creating a Password and Adding Credit

    menu lateral direito - view API keys

    Go to https://platform.openai.com/ and click on API. After logging in, in the upper-right corner, click “View API keys,” as shown in the adjacent image. Then, in the left-side menu, go to “Create new secret key.”

    🚨 The API only works if you add credit to your account.

    To add credit, go to Billing. The amount will be charged in dollars to your card, so I recommend using a global card with lower fees, like Nomad, which I discuss in my travel posts.

    How does the cost work?

    The plugin consumes tokens based on the number of processed words. So, every time you “chat with your notes,” as seen above, it counts the characters in the response as processed words. Depending on your usage, costs may vary, but they are generally affordable, and more importantly:

    🚨 You can set usage limits to avoid surprises on your bill.

    To set spending limits, go to Rate limits. I set a limit of $1 per month, which I found sufficient for around 100 searches in my notes. When you reach half of your monthly usage, you receive an email alert, which I found very secure.

    email que recebi quando cheguei a metade do uso dos créditos do mies
    email que recebi quando cheguei a metade do uso dos créditos do mies

    Custom Settings

    After entering your API key, you can test it, but it will only work if you have credit.

    Additionally, the plugin offers various customizable settings, such as language model selection, file exclusions, and display options. You can tailor the plugin to your specific needs for a personalized experience.

    • Language Model: I use the “3.5 Turbo 16K” model for fast performance. However, it also offers the GPT-4 model.
    • Exclusions: You can specify which files or pages to exclude from the plugin’s processing. I included the plugin’s own created file in the exclusion list.
    • Display: You can choose to have an expanded view of all notes, but I prefer the simple view with just the current note and the plugin chat open.

    You can adjust these settings according to your preferences and requirements. Below is a screenshot of my settings:

      menu de configurações do plugin no obsidan
      menu de configurações do plugin no obsidan

      Finally, after testing Notion’s AI, ChatGPT, and Smart Connections from OpenAI, which is a different product from ChatGPT, I dare say I prefer Smart Connections, both in terms of cost-effectiveness and the quality of generated content. Of course, this is only possible when you have notes you can trust.

      I hope this helps! 


      Pin this post