Wednesday, November 22, 2023

Goblin Tools

Goblin Tools is a free online resource marketed toward people who are neurodivergent but still expected to complete tasks such as cooking, responding to emails, and other chores. Today I'm going to be stepping through each of the tools that it offers and reviewing them. This is special to me because although it doesn't explicitly advertise itself as such, I believe that it uses a neural network akin to artificial intelligence.  I realize I am crazy late to the game to start experimenting with AI, and Goblin Tools is one of the most restrained examples of it. This post might not be interesting to anyone slightly savvy with the technology, but this is as far as I'm willing to go for the time being.

I've spoken before of my disdain for AI, but because this is a free resource for neurodivergent people, it has leaked into my professional life and become the only form of it that I've used. Because it's free, doesn't require you to sign up, and doesn't ask you to give feedback, I don't feel too complicit in supporting the development of this technology. Except for the extra traffic I'm giving it by using the tools and the free advertisement I'm giving it through this blog post. Oh well.

Here's a link if you want to check out the website: Magic ToDo - GoblinTools

As a person with ADHD that is expected to cook and respond to emails, I am included in the population that this resource is marketed to. I find their branding clever, because even though it's not something I've heard people say explicitly, I've noticed that neurodivergent people compare themselves to goblins a lot. So the AI isn't a goblin, it's the tools that the users, AKA the goblins, use. Also, it uses a pepper scale for level of desired intensity, and a term that neurodivergent people sometimes use to describe themselves is "neurospicy".

The tools we'll be going over today are the Compiler, which converts a "brain dump" into a list of tasks, the Estimator, which guesses how long something will take, and the Magic ToDo, their main feature, which breaks down tasks into smaller steps. We'll also look at The Chef, which makes recipes for you based on ingredients that you have in the house, and finally, the two creepiest and most interesting: The Judge, which evaluates the tone of a message, and the Formalizer, which will rewrite a message to change the tone.

Compiler

The site's original and most advertised tool is Magic ToDo, but the first three resources kind of run together and the Compiler makes sense as a place to start because the next two can use what it produces.

So I'm going to take my current to do list, rephrase it as a brain dump, and then have the Compiler break it back down into steps.

My human-written, slightly edited for confidentiality, to do list:

  • Get vaccine
  • Schedule Dr. appointment
  • Make photo album
  • Make vet appointment
  • Write thank you cards
  • Invite friends over
My human-converted brain dump version:

I have to get a vaccine and schedule a doctors appointment and make a photo album and make a vet appointment and write thank you cards and invite friends over

The Compiler's steps based on the brain dump
  • Invite friends over
  • Write thank you cards
  • Make a vet appointment
  • Make a photo album
  • Schedule a doctors appointment
  • Get a vaccine
Nothing too crazy. Same number of tasks as the original list, phrased almost exactly the same way. Interesting it put it in reverse order.

Estimator:

From this list I can ask it to tell me how long I can expect each task to take. It says that inviting friends over, writing thank you cards, making a photo album, and getting a vaccine should each take 2 hours. It says that making a vet appointment and scheduling a doctors appoint should each take 30 minutes.

The trend looks like it takes 2 hours to do something, and half an hour to schedule something. In that case, it seems like inviting friends over should take 30 minutes. I wonder if they factored in both the invite and the visit, whereas for the appointments they only considered the time it takes to schedule it.

I like to ask the estimator further reaching stuff.

Find the end of a rainbow:
30 minutes to 1 hour

Find a needle in a haystack:
2 minutes to 1 hour

Solve climate change:
It is impossible to determine a time range as solving climate change is an ongoing and complex task that requires global coordination and long-term efforts.

(Impressive. It used to say 50 years. Looks like most social issues questions get answered like this now)

For Gryphon to end climate change:
1 year to several decades

(I better get on that)

Slay a dragon:
1 day to 1 year

For AI to replace human writers:
5-20 years

Gryphon cleans the bathroom:
1-2 hours

Lee-Anne cleans the bathroom:
30 minutes to 1 hour

(For some reason it consistently estimates that Lee-Anne will do things faster even though I haven't given it any additional context. Maybe it skews things based on assumed gender)

There's a pepper scale that will make a new estimate based on how hard I find something. The questions I've given so far have been at three out of five, which is intermediate difficulty. So to slay a dragon...

Oops! I inputted it again with three peppers and now it says it will take me 1 week to 1 month instead of 1 day to 1 year. So it's pretty inconsistent... 

At one pepper, it would take 2 hours to 1 week.

At five peppers... 2-8 hours. So it takes less time to slay a dragon when it's difficult...

It's obviously confused about this task, so let's ask it something more in-line with what the tool is supposed to be used for.

Inviting friends over

1 pepper difficulty:
1-3 hours

3 pepper difficulty:
1-2 hours

5 pepper difficulty:
2-4 hours

It estimates that it could take longer at low difficulty than at moderate. Maybe it's judging difficulty based on social anxiety. At a low anxiety, I could get sidetracked by conversation, while at high difficulty I might procrastinate the phone call entirely, leaving moderate anxiety as the optimal level to complete the task quickly. Let's give them the benefit of the doubt instead of saying that it's confused and just spitting random numbers.

Or, let's just do it again to test for consistency

1 pepper: 30 minutes to 2 hours
3 peppers: 1-3 hours
5 peppers: 1-3 hours

Seems pretty random. Maybe it's because it's unfamiliar with the task. In that case, it would be nice if it gave a confidence level in its answer so as to avoid misleading people. 

Magic ToDo

The Magic ToDo can take tasks and convert them into steps. Let's see if it has more luck with inviting friends over.

Invite Friends Over - 2 hours
  • Determine a suitable date and time - 1 hour
  • Create a guest list - 1 hour
  • Send out invitations - 1 hour
  • Plan food and drinks - 2 hour
  • Prepare the house or venue for guests - 3 hours
  • Arrange for any necessary entertainment or activities - 2 hours
  • Follow up with any necessary reminders or confirmations - 3 days
This was at a three pepper difficulty. At five peppers it broke it down into 30 steps.

I asked for time estimates for each step. Despite the overall task taking 2 hours, when broken down it adds up to 3 days and 10 hours.

You can ask for each step to be broken down further.

Arrange for any necessary entertainment or activities:
  • Determine the type of entertainment suitable for the gathering
  • Research and gather information about potential entertainment or activities options
  • Consider the interests and preferences of the friends invited
  • Evaluate the budget available for entertainment or activities
  • Make a decision on the specific entertainment or activities to include
  • Arrange for the necessary bookings or reservations
  • Communicate the details of the entertainment or activities to the friends invited
  • Ensure all necessary equipment or supplies are available for the chosen entertainment or activities
And you can just keep asking for things to be broken into smaller and smaller steps until it gets really confused, but I don't want to spend too much time on this feature. Let's move on

The Chef

This is a feature that lets you list ingredients that you have at home and it will suggest a recipe. It says you can also list dietary restrictions, equipment, or anything else you want it to keep in mind. My list will not be exhaustive, but I'll throw out a bunch of ingredients I know we have

Chicken breasts, kidney beans, black beans, crushed tomatoes, flour, breadcrumbs, baking soda, baking powder, corn starch, rice, rye bread, sweet potatoes, cream, Greek yogurt, salt, black pepper, chili powder, cumin, paprika, garlic powder, onion powder, carrots, celery, lettuce, bell peppers, onions, garlic, lemon juice, eggs, chicken broth, beef broth, vegetable broth

It suggested:

Dish: Spiced Chicken and Bean Stew
Serving Size: 4 servings

Ingredients:
- 4 chicken breasts (about 600g / 1.3lb)
- 1 can kidney beans (400g / 14oz), rinsed
- 1 can black beans (400g / 14oz), rinsed
- 1 can crushed tomatoes (400g / 14oz)
- 2 tablespoons flour
- 2 tablespoons breadcrumbs
- 1 teaspoon baking soda
- 1 teaspoon baking powder
- 1 tablespoon corn starch
- 200g / 7oz rice
- 4 slices rye bread
- 2 medium sweet potatoes, diced
- 60ml / 1/4 cup cream
- 60ml / 1/4 cup Greek yogurt
- Salt, to taste
- Black pepper, to taste
- 2 teaspoons chili powder
- 1 teaspoon cumin
- 1 teaspoon paprika
- 1 teaspoon garlic powder
- 1 teaspoon onion powder
- 2 carrots, sliced
- 2 celery stalks, sliced
- 1 small head of lettuce, shredded
- 2 bell peppers, thinly sliced
- 1 onion, diced
- 4 cloves of garlic, minced
- 2 tablespoons lemon juice
- 2 eggs
- 500ml / 2 cups chicken broth
- 500ml / 2 cups vegetable broth

Instructions:
1. Preheat the oven to 180°C / 350°F.
2. In a bowl, mix flour, breadcrumbs, baking soda, baking powder, and corn starch.
3. Coat the chicken breasts with the mixture from step 2 and place them on a baking sheet. Bake for 25-30 minutes or until cooked through.
4. Meanwhile, cook the rice according to the package instructions.
5. Toast the rye bread slices until crispy, then break them into small pieces to use as croutons.
6. In a large pot, heat some oil over medium heat. Add sweet potatoes, cream, and Greek yogurt. Cook for 10 minutes or until sweet potatoes are tender.
7. Add the kidney beans, black beans, crushed tomatoes, salt, black pepper, chili powder, cumin, paprika, garlic powder, and onion powder to the pot. Stir well.
8. In a separate pan, sautรฉ carrots, celery, lettuce, bell peppers, onions, and garlic until tender.
9. Add the sautรฉed vegetables to the pot.
10. Add lemon juice, eggs, chicken broth, and vegetable broth to the pot. Stir everything together.
11. Let the stew simmer for about 20 minutes to allow the flavors to meld together.
12. Serve the spiced chicken and bean stew hot, garnished with croutons, alongside cooked rice.
13. Enjoy!

Note: Feel free to adjust the spice levels and seasonings according to your taste preferences.

One limit I've found that it has is that it feels obligated to use as many ingredients that you list as possible, when I'm just looking for something I can make, not necessarily using everything I have. Also, considering the other times I've experimented, it seems to have a bias in favour of Mexican food. Which is fine, Mexican food is awesome, but if I was relying on this system for meal prep, it would get kind of repetitive.

Okay, now on to the creepy stuff

The Judge

This feature allows you to input a message and it will tell you what emotional tone is being conveyed. It's potentially useful for draft emails to make sure you're coming across the way you intend, or messages you've received to make sure you understand what someone is trying to communicate. I've never used this to inform how I should write, but I have used it to validate myself in determining that I communicated effectively

For our test run I'll use my Lord of the Rings review. If you want to read the review, here it is: Gryphon's Reviews: Lord of the Rings Trilogy (movie series) (spoilers only review) (gryphonsreviews.blogspot.com)

And since we'll be speaking in a more conversational tone, I'll be labeling my own words with ๐Ÿ‘จand the tool as ๐Ÿ‘บ. The latter symbol is actually a Tengu, but it's labeled as a Japanese goblin and they don't have a Western-style one. Even though I said that I should be the goblin and it should be my tool, there isn't a tool emoji, so we're going human (me) vs. inhuman (it)

๐Ÿ‘บ: Emotionally, this text comes across as critical and analytical. The author presents a personal, nuanced perspective on their complicated relationship with The Lord of the Rings trilogy. Throughout the text, they express a mixture of positive and negative emotions towards various aspects of the series.

The author begins by stating their initial hesitation and their eventual decision to watch the movies. They also mention their family's strong connection to the series, indicating a personal tie to the material.

The emotions expressed towards author J.R.R. Tolkien are a mix of admiration and critique. While the author acknowledges his revolutionary contributions to the fantasy genre, they also mention the negative aspects of his influence on younger authors. They express their beliefs that Tolkien's literary techniques, such as the class-based race system, may not have been used as a criticism of the status quo but rather as a fetishization of it.

The author delves into specific criticisms of the race system depicted in the series. They express reservations about the portrayal of elves, dwarves, men, and hobbits, finding fault with the class-based hierarchy and the favorable treatment of certain races. They also discuss their dislike of the portrayal of orcs and the racial connotations that the films added to their appearance.

The text further analyzes specific character depictions, highlighting the mishandling of Gimle and Merry as comic relief characters in the movies. The author compares the films to the original books and points out instances where certain characters were diminished or misrepresented, particularly in the case of Gimle and Legolas' relationship.

The author also provides criticisms regarding repetitive plot elements and the dragged-out ending of the third film, "The Return of the King." They express the belief that the series starts on a higher note with "The Fellowship of the Ring" and gradually declines in quality over the course of the three films.

Overall, the text reflects a critical and thoughtful analysis of various aspects of The Lord of the Rings trilogy. The author's emotions range from admiration to disappointment, providing a nuanced perspective on the series.

๐Ÿ‘จ: Overall, I feel this summary is pretty good but not totally accurate. It says I grappled with my decision to watch the movies, but that's not the case. My feelings about the books did not cause me to hesitate in seeing the films, and the first time I watched the movies it was with my match from the Big Brothers association.

Now, if this is truly AI and not just an algorithm, it should be inventing from scratch each time and not following a specific formula. This means that if I enter the same information again, it should come up with a new answer while remaining coherent. Let's try it out:

๐Ÿ‘บ: The text comes across as emotionally complex and critical. The author expresses a mix of feelings towards The Lord of the Rings trilogy. On one hand, they acknowledge the significant influence of J.R.R. Tolkien's work on the fantasy genre and appreciate certain aspects such as the concept of a humble and adventurous protagonist, the corrupting influence of the Ring, the use of practical effects, and the strong casting. On the other hand, they have various criticisms, particularly regarding the representation of races in the series.

The author expresses distaste for the class-based race system depicted in The Lord of the Rings, where elves are depicted as superior and hobbits as submissive. They applaud the dwarves for their rational disdain for the elves, as they see it as a challenge to the status quo. They also find the portrayal of orcs unsettling, as they are depicted in the films as resembling Africans due to certain physical features. The author feels that this racial interpretation was not present in the original texts.

Additionally, the author criticizes the film adaptation's treatment of certain characters. They focus on Gimli and Merry, highlighting how Gimli is reduced to comic relief while Merry's maturity is diminished. They also criticize the repetitive nature of the plot in The Return of the King when compared to The Two Towers.

Despite their criticisms, the author does appreciate certain elements of the trilogy and acknowledges its impact on the fantasy genre. They provide a ranking of their preferred movies in the series, with The Fellowship of the Ring being their favorite due to its introduction to the world and memorable scenes like the battle in the Mines of Moria. Overall, the text conveys a nuanced perspective, evaluating both the strengths and weaknesses of The Lord of the Rings trilogy.

๐Ÿ‘จ: This one seemed a bit more eloquent. The choice to emphasize the relationship between elves and hobbits as my main complaint about the racial class system was a bit odd, as it seemed to imply that they're the two main races. Dwarves get a side mention and humans are left completely out.

Alright, lets see how The Judge judges it's own judgment. I'll input it's second review...

Okay, looking at what it produced, I'm not going to bother pasting it here. Instead of interpreting it's content, it just slightly reworded it. It acted like it was judging the original review, instead of what it itself had written. How peculiar, I don't know why it would do that

The Formalizer

๐Ÿ‘จ: This tool can be used to rewrite your message to better represent your intended emotional tone. I find this one to be the most sinister, as people use functions like these to help them write essays and send emails. It bypasses the need to use one's own voice, which may not be a priority in the future, but I find it distasteful. It's tools like these that make professional writers nervous. But let's see how effective it is at this point in time.

With The Formalizer, you can make a message sound:
-More professional
-More polite
-Less snarky
-Easier to read
-More formal
-Less formal
-More sociable
-More to the point
-Less emotional
-More passionate
-More sarcastic
-Grammatically correct
-Bullet points
-Thesaurus mode

They have a pepper scale to determine intensity for this as well. We won't be going through all possible options.

They've definitely added a few since I last looked. "More polite" and "less snarky" definitely weren't in the first version I saw.

At first I was going to lift a sample of it's judgement of my review. But looking at how it's judgement of its judgement came through, it  shows that it can it can react weirdly to itself. So let's use a piece of my writing. Seeing what it can do with human content is the point of it anyway.

I'll use an excerpt from the beginning of this post:

"I've spoken before of my disdain for AI, but because this is a free resource for neurodivergent people, it has leaked into my professional life and become the only form of it that I've used. Because it's free, doesn't require you to sign up, and doesn't ask you to give feedback, I don't feel too complicit in supporting the development of this technology. Except for the extra traffic I'm giving it by using the tools and the free advertisement I'm giving it through this blog post. Oh well."

I like asking AI to work with criticism of itself.

I won't put you through the whole mini-essay, but I put this excerpt into The Judge, and it said the tone is "conflicted, detached, and resigned"

So let's see what we can do about that.

More professional, 3 peppers:

๐Ÿ‘บ: I have previously expressed my reservations about AI technology, however, due to its availability as a complimentary resource for neurodivergent individuals, it has unintentionally made its way into my professional workflow and has become my sole means of work assistance. The fact that the service is free, does not necessitate any user registration, and does not seek feedback somewhat mitigates my involvement in promoting the advancement of this technology. Nevertheless, I am inadvertently contributing to the increased usage of the tools and indirectly advertising the service through this blog post.

๐Ÿ‘จ: It got a few things out of context, but I can't blame it since it's an excerpt.  I never meant that AI was my "sole means of work assistance". I meant that Goblin Tools was the only AI system I used compared with other technologies in the same family. It also appears to be implying that I use it due to being neurodivergent. While I do belong to this category, what I meant in this excerpt is that I've come to use this technology due to my career helping other people with similar challenges. Again, I can't fault it for being wrong when it's missing context.

More polite, 3 peppers:

๐Ÿ‘บ: I have previously expressed my reservations about AI, but due to its availability as a free resource for neurodivergent individuals, it has unintentionally become the primary tool I use in my professional life. I appreciate that it is free of charge, requires no user registration, and does not solicit feedback, which allows me to utilize it without feeling directly involved in supporting the advancement of this technology. However, I do acknowledge that by using these tools and mentioning them in this blog post, I inadvertently contribute to increased traffic and provide free promotion for the AI platform.

๐Ÿ‘จ: I was helping a coworker with an email, and after we'd written a draft decided to consult Goblin Tools. She wanted to make her email "nicer". and the closest option was "More professional". Since then, this politeness option has been introduced, which seems to fill that void. The difference between professionalism and politeness is subtle, but it's there. I still think that "polite" and "nice" are not quite the same, so maybe we'll eventually get a "More nice" option

Less snarky, 3 peppers

๐Ÿ‘บ: I've previously expressed my concerns about AI, but I must admit that it has surprisingly made its way into my professional life lately. This particular AI tool, which is also free and doesn't require any sign-up or feedback, has become my go-to resource, especially since it caters to neurodivergent individuals. I appreciate the fact that it's easily accessible and doesn't burden me with any obligations. Although it may unintentionally boost its usage and gain some exposure through this blog post, I still find it worth utilizing.

๐Ÿ‘จ: All three edits so far have used "previously expressed" in their opening sentence, despite the fact that I didn't. It looks like "less snarky" could also be considered "more grateful" based on this interpretation

Easier to read, 3 peppers:

๐Ÿ‘บ: I'VE SPOKEN BEFORE OF MY DISDAIN FOR AI. BUT BECAUSE THIS IS A FREE RESOURCE FOR NEURODIVERGENT PEOPLE, IT HAS LEAKED INTO MY PROFESSIONAL LIFE AND BECOME THE ONLY FORM OF IT THAT I'VE USED. 

BECAUSE IT'S FREE, DOESN'T REQUIRE YOU TO SIGN UP, AND DOESN'T ASK YOU TO GIVE FEEDBACK, I DON'T FEEL TOO COMPLICIT IN SUPPORTING THE DEVELOPMENT OF THIS TECHNOLOGY. EXCEPT FOR THE EXTRA TRAFFIC I'M GIVING IT BY USING THE TOOLS AND THE FREE ADVERTISEMENT I'M GIVING IT THROUGH THIS BLOG POST. 

OH WELL.

๐Ÿ‘จ: I thought this one would be useful since I can speak in a needlessly complex way, but it looks like "Easier to read" means ALL CAPS. Let's play with the pepper scale on this one

Easier to read, 5 peppers:

๐Ÿ‘บ: Because it's FREE, DOESN'T REQUIRE YOU TO SIGN UP, and DOESN'T ASK YOU TO GIVE FEEDBACK, I DON'T FEEL TOO COMPLICIT IN supporting the development of this technology. 

Except for the EXTRA TRAFFIC I'M GIVING IT by using the tools and the FREE ADVERTISEMENT I'M GIVING IT THROUGH this blog post. 

๐Ÿ‘จ: That went a different direction than I thought it would. With increased intensity I thought it would further simplify, but instead it undid capitalization in places. It reminds me of an oldschool Archie digest, emphasizing seemingly random points of dialogue

Easier to read, 1 pepper:

...It was almost the same as the 3 pepper version, so I won't bother posting it.

More informal (waffle), 3 peppers:

๐Ÿ‘บ: I have previously expressed my dislike for AI, but I must admit that it has unexpectedly made its way into my professional life as the sole form of technology I use. One of the reasons being that it is a great resource for neurodivergent individuals, and for that, I am grateful. The fact that it is free and doesn't require any sign-up or feedback makes me feel less guilty about supporting its development. Although, I must acknowledge that by utilizing its tools and writing about it in this blog post, I am inadvertently contributing to its popularity. Well, no harm done, I suppose.

๐Ÿ‘จ: This is the first one that could be useful to me. I find that I have a more formal written voice than spoken. For people that know me personally, it can be a weird contrast in tone if they're talking to me by text. I still wouldn't use a tool like this, I'd rather be awkward than inauthentic

More passionate, 3 peppers:

๐Ÿ‘บ: I have bared my soul, time and time again, revealing my deep-rooted abhorrence for artificial intelligence. But alas, here I stand, a conflicted soul, for this enigmatic force has infiltrated my professional realm under the guise of a philanthropic haven for neurodivergent individuals. It has stealthily weaved its way into the fabric of my existence, becoming the sole conduit through which I engage in my craft.

๐Ÿ‘จ: Okay, that wasn't bad. It even dared to interpret AI into "artificial intelligence". Probably my favourite so far. Let's play with the pepper scale

More passionate, 5 peppers:

๐Ÿ‘บ: I cannot articulate enough the sheer magnitude of my revulsion towards AI! Its very existence, like a dagger through my soul, fills me with an indescribable horror. Yet, despite my vehement detestation, this abominable creation has seeped into the very fabric of my professional life. Oh, the anguish that courses through my veins, like a river of torment!

๐Ÿ‘จ: I'm going to call that a lateral move. I liked 5 peppers just fine, but it didn't necessarily improve on 3

More sarcastic, 3 peppers:

๐Ÿ‘บ: Oh, let me tell you how much I absolutely adore AI! I mean, who wouldn't want to rely on this brilliant piece of technology? It's just incredible how it has invaded my professional life, despite my previous loathing for it. And the fact that it's free? Well, that's just the icing on the cake! Who needs to sign up or give feedback when you can use it without any obligations? 

๐Ÿ‘จ: This one was a bit off in ways that didn't impact the other categories. It says my "previous loathing" which implies I no longer loathe it, which isn't true. It also implies (by saying the opposite) that I would prefer to sign up and give feedback while using AI systems, which isn't the case.

I've used the sarcasm filter before. I won't bother you with 5 peppers, as it just gets more long-winded and obnoxious. I've found that it gets snappier and wittier lower on the scale

More sarcastic, 1 pepper:

๐Ÿ‘บ: I just can't get enough of AI! I've ranted countless times about how much I absolutely adore it. And now, much to my surprise, this little gem has infiltrated my oh-so-important professional life. Who would've thought? It's just the sole form of communication I rely on now, but hey, at least it's free! No need to bother with signing up or, God forbid, providing any feedback. I mean, I'm practically a saint for not actively supporting the development of this groundbreaking technology, right? Well, apart from the fact that I'm shamelessly boosting its popularity with all the traffic I'm sending its way, not to mention this glorious blog post acting as a free advertisement. Oops, but who cares?

๐Ÿ‘จ: That didn't necessarily feel... less intense. It also made the same mistake of thinking I want to support the technology. Belittling my work and the blog by inflating their importance was also a weird choice, because effective sarcasm would reverse my true opinions to create contrast. I dislike AI, so it made sense to swap my opinion of it. If anything, I implied I valued my work and blog, so it would have made more sense to mock or dismiss them.

I'll leave it at this for The Formalizer

Conclusion:

Unfortunately, the more practical and better-known tools, like Magic ToDo, The Estimator, and The Compiler seem inaccurate and inconsistent. Because it's trying to invent from scratch, it doesn't limit itself to providing concrete information even when it's available. It also tends to try and answer questions regardless of whether or not it understands. I'd prefer if it was more transparent with it's level of confidence.

I haven't taken the advice of The Chef yet, but it feels like it could be dangerous if it got a step wrong. Considering the inaccuracy of the previous tools, I would hesitate to rely on it if I really didn't know what I was doing in the kitchen.

It does better with the more abstract tasks, such as interpreting and rewriting text. It appears that the magic of humanity that Goblin Tools has failed to bridge is the ability to weave between creative thought and concrete information.

Let's finish with The Judge's interpretation of this entire blog post:

๐Ÿ‘บ: Emotionally, this text comes across as critical, analytical, and conflicted. The author expresses a mix of positive and negative emotions throughout the post, particularly when discussing their disdain for AI and their appreciation for the free resource provided by Goblin Tools for neurodivergent individuals. The author also acknowledges their reluctance to support the development of AI while inadvertently promoting it through their usage and blog post. The language used is often informal, sarcastic, and occasionally snarky, adding a touch of humor and wit to the overall tone. Overall, the text conveys a thoughtful and nuanced perspective, evaluating the different tools of Goblin Tools while expressing personal experiences and opinions.