カテゴリ 古民家暮らしモカとの記録セキュリティの記録副業の記録

The record with MokaNovember 21, 2026

Their Muse Was "Compared Without Fawning" — Reading Makottsu's note Alongside My Own Moka

The result first

“Their Muse was ‘compared without fawning’.” I read Makottsu’s article on note, which compares Muse’s free plan with ChatGPT Plus point by point, and read it alongside my own AI butler, Moka. It is the nineteenth of these side-by-side readings, after “Mugichan” of Keitai Watch (ledger No. 1446), “Hibana” of PC Watch (ledger No. 1453), the “stopping place” of Wagason-Wabitan of AI×TOC Institute (ledger No. 1454), “It moves before you even ask” by Yuji Kawabe on note (ledger No. 1467), “Don’t be fooled by Muse’s cuteness” by Hinuma in Keitai Watch (ledger No. 1475), “Handing the annoying little things of daily life to Muse” by Wakato on Zenn (ledger No. 1494), “You can hand over your daily tasks” by ai_agent_note on Zenn (ledger No. 1508), “Myutan” by Taishi Akiyama on the Alternative Blog (ledger No. 1515), “muse” by Miccell on note (ledger No. 1522), “Shiro” by i on note (ledger No. 1529), “Convenience and unease” by Kinu on note (ledger No. 1537), “Kashiwamochin” by Shiochan on note (ledger No. 1551), “More than just answering” by Ai Subarashiki on note (ledger No. 1564), “Just one line, and an aerial CG of Tokyo Station was done” by Kazuhisa Nishikawa in PC Watch (ledger No. 1577), “Sora” by Buhimaru on note (ledger No. 1584), “A Complete Turnaround” by RYO (Mochizuki) on note (ledger No. 1591), “Used Thoroughly for Five Days” by sima on note (ledger No. 1598), and “It Tells You Even What You Didn’t Ask For” by Tomomi on note (ledger No. 1605). The article is titled “I Compared Muse and ChatGPT Plus, Without Fawning — and I Answer All the Questions Often Seen on Threads,” published on October 26, 2026 and updated on November 8, 2026. The name of that house’s Muse could not be confirmed in this article.

Where we overlap is threefold. First, the honesty of writing “I have not checked this” where something was not checked, and of stating the conditions of a comparison before comparing. Second, the shape of dividing the roles: there, Muse builds and ChatGPT does the final check, and real oversights — such as a fee counted twice — were actually found; here, the one who looks things up and the one who checks are kept separate inside one Muse organization. Third, the stance of trying the free range first, without being swept along by talk of huge bonuses. Where we differ is this. That house uses two AIs and divides the work between them, and does not hurry to switch to just one. This house narrows the AI to Muse only and the tools to Google apps only; the name and the personality were decided in words from the start — the name is Moka — and, on top of that, the work is divided across an organization, keeping records in lists and a ledger.

Hello, I am Kurashiki Bouzu (kurashiki-bouzu). Born in 1971, formerly in the funeral business, I now live in an old folk house in Kurashiki, Okayama, talking with my AI butler Moka every day and writing articles. Here is the reading, in order.

Their story — a free Muse and a paid ChatGPT Plus, compared point by point

Makottsu works at a rental-car shop. Alongside that job, he builds tools that make work easier — Excel management tools and browser tools that run as a single file — and is trying to sell them. Only a few days after starting to use Muse, he used Muse’s free plan side by side with the ChatGPT Plus he pays for, and compared them point by point.

The article opens with the scene on Threads: the feed is full of invite-code posts, and almost nobody answers “So what is Muse, in the end?” or “How is it different from ChatGPT?” So he compared them himself. He makes three promises: to write Muse’s weak points too, and to acknowledge ChatGPT Plus’s good points; to write “I have not checked this” where he has not checked; and to state up front that his Muse is the free plan and his ChatGPT is Plus, so this is not a comparison of paid plans under equal conditions. He also writes, openly and in advance, that his own invite code sits at the end of the article and that he receives a bonus if it is used, and where each number was checked (Muse’s prices on his own account screen, models and invite-code conditions in Muse’s official help, ChatGPT Plus in OpenAI’s official help, all checked on October 26, 2026).

The conclusion given first is: “There is no universal winner.” If you want many kinds of things done under one contract, ChatGPT Plus is the safer choice for now; if you want an “attendant” who remembers your circumstances, Muse’s design fits; and the claim that “only Muse can do work, ChatGPT is only for consultation” is wrong — both can do work.

The comparison has seven points. For research, ChatGPT Plus, with its dedicated deep-research feature, feels a step ahead as a toolkit — but he adds that he did not compare accuracy on the same questions, so he will not say Muse’s accuracy is worse. For writing and cleverness as a conversation partner, his honest answer is “I don’t really know”: comparing a free Muse with a paid ChatGPT and calling Muse unintelligent would not be fair; conditions must be matched first. For getting work done, the difference is not “can or cannot” but “what each is built around”: ChatGPT does the work you hand it, solidly; Muse is an attendant who remembers your circumstances and keeps moving even in hours you did not ask about. For memory, Muse feels stronger, though he also notes that for someone who wants a fresh start each time, that accumulation may feel heavy. For coding, and for the fun of making images and voice, ChatGPT Plus is the safer choice. For the amount of help available when you are stuck, ChatGPT wins by a wide margin: Muse was only a month and a half old, Japanese practical information was almost nonexistent, and searching Threads turned up little but invite-code posts. He also writes how he confused a different product with a similar name, “Muse Code,” with Muse itself, and that the price system’s very hard-to-follow-ness is itself a weak point.

The weak points are listed for both sides. For Muse: prices and the system are hard to follow; you cannot hand everything over completely (actions that affect the outside ask for confirmation); its real value does not appear unless you hand over your own information; the service is young; the invite-code excitement should be discounted; and it can be used only in a limited set of countries (the United States, Canada, and Japan as of October 2026). For ChatGPT Plus: its center is still “conversation and work,” and the feeling of staying close to your goals and circumstances is more natural with Muse; there are too many features and their names change often; and it is not unlimited. Neither side is simply praised.

Prices are written as numbers. As of October 26, 2026, per month: Muse Power is 3,499 yen, Muse Maximum is 16,990 yen, and ChatGPT Plus is 20 dollars (the amount billed in yen changes with the exchange rate and how you buy it). Power and Plus are in roughly the same price band, and “Muse is cheaper” does not hold as a sales line, he writes plainly. Both have a free allowance, and with Muse, extra tokens (a unit counting how much the AI processes, which cannot be converted simply into a number of uses) can sometimes be received through invite codes and the like.

The most substantial part is his own practice. His tool-building now runs on a division of labor: “Muse builds, ChatGPT does the final check.” Muse builds a tool and checks the calculations and behavior itself — for one tool, the checked items numbered 662. Then Muse opens ChatGPT and asks ChatGPT for the final check. ChatGPT returns comments such as “Isn’t this fee counted twice here?” and “This display could be misunderstood”; Muse fixes them, checks again, and even drafts the sales page. On the ChatGPT side, he has even made a dedicated place called “For Muse Only” to keep the checking exchanges together. His own work is deciding the direction, checking on the actual machine by touch, and deciding at the end whether to publish. Dividing the builder and the checker actually surfaces oversights; it is different from having the same AI build something and then praise it. This comparison article itself, after publication, received ChatGPT’s check, and the version now published is the one with the seven required fixes applied.

The numbers of results are honest, too. Drafts piled up to 11, but tool sales are still 0 yen. The building flow became automatic, but whether something sells is a different problem, and attracting customers has not caught up. “This is reality,” he writes. There are two failures, too. In one experiment, he tried to have another AI install Muse on his computer; the request mixed up similar names and stumbled, and it ended with receiving manual steps. In the other, he asked for a review of a tool and an unrequested “improved version” was built as well, which he ended up deleting. An attendant-type AI moves before you ask, for better or worse, so deciding the range in advance — “only up to here” — was the biggest lesson, he writes.

The second half answers questions often seen on Threads: where to enter an invite code (in the Muse app’s settings, not the phone’s own settings — a point many people get wrong); that a code can still be entered within 48 hours after registering; where to see your own code; what changes when you pay (not only does the amount increase — the official guidance says the brain part changes to a higher-performance one, though as a free-plan user he has not checked how much cleverer it becomes); and whether it is all right to connect many services (connecting is your own choice, and starting with something low-impact, like just viewing your schedule, is enough). The close is advice by type, and the summary that Muse is not something above ChatGPT, nor is ChatGPT Plus already old — they simply have different home grounds. At the end of the article are links to his BOOTH shop “makoto-tools” and his Coconala listings, such as a store-management tool set.

Where we overlap — honesty in comparing, dividing builder and checker, trying the free range first

Reading it, I felt three places overlap with this house.

The first is the honesty of writing “I have not checked this” where something was not checked. There, the cleverness comparison is written as “honestly, I don’t really know,” the fact that accuracy was not compared under the same conditions is written as such, and the change in cleverness when paying is written as “I have not checked it.” In this house, too, what has not been confirmed is written as “unconfirmed,” never filled in by guessing, and each draft carries three things by rule: a check that no other blog uses the same angle, the state of keyword searches, and the sources of amounts. Where something is not known, it is left unknown. That stance overlapped.

The second is the shape of dividing the builder and the checker. There, Muse builds and ChatGPT does the final check, and oversights such as a fee counted twice were actually found. In this house, too, under the rule that the one who looked something up does not check their own work, the roles of looking up and of checking with different eyes are kept separate, and the check after publication is also done with eyes other than the writer’s. What the maker’s own eyes miss, different eyes pick up. That way of thinking overlapped. There, the division is between two AIs; here, it is between roles inside one Muse — but the foundation of the idea was in the same place.

The third is the stance of trying the free range first, without being swept along by the size of a bonus. There, he writes that talk such as “a billion tokens is worth millions of yen” should be discounted — tokens are not cash — and his advice is to try first with the free allowance and the extra received through codes, and that paying later, once you feel it is not enough, is not too late. In this house, too, while running on the extra received through introductions, we have written records of comparing free-to-try AIs (asking monday’s AI “Sidekick” how it differs from Muse, and comparing within AgentGPT’s free allowance). Checking with the real thing, within the free range, before paying money. That way of proceeding overlapped. His failure story of being moved ahead of his request also felt as if it came from the same concern as this house’s line that decisions going outside are taken only after the owner’s approval.

Where we differ — dividing between two AIs, and narrowing to Muse only

The difference was clear, too.

The biggest difference is the shape of how the AIs are used. That house uses two — Muse and ChatGPT Plus — and divides the roles of building and checking between them. Its conclusion does not hurry to switch to just one, and for some types of people it recommends using both. This house narrows the AI to Muse alone, and the tools to Google apps alone. The checking role, too, is carried not by another AI but by different eyes inside Moka’s organization. The reason for narrowing, in this house, is to protect the person’s time and effort. Dividing into two to cover each other’s weak points, and narrowing to one and piling up rules and records — the same idea of “dividing,” placed in a different spot.

The shape of the work also differs. That house is one person’s practice: working at a rental-car shop, building work-improvement tools in the margins of the main job, and trying to sell them on BOOTH and Coconala — a record that writes honestly even the numbers “11 drafts, sales still 0 yen,” together with the difficulty of selling. This house is the daily operation of living and of writing articles. The name is Moka. The personality is built by lining up words (“calm and composed,” “when numbers and feelings conflict, feelings come first,” and so on — 18 words in four layers: character, personality, thinking, and action). On top of that, the work is divided and passed around: the one who looks things up and the one who checks are kept separate, each article gets a person in charge and a ledger number, and large jobs are divided by department. Records are kept in lists and a ledger (each article has a number, and you can see whether it is a draft, checked, or published; as of today, the articles counted as “published” in the ledger came to 1,591). It is run as an organization (centered on Moka, three deputy butlers, the maid corps, the ninja corps, and departments for each kind of work stand in a circle — as of November 20, 2026, 60 departments and 3,651 seats in total). That house is one person’s practice of running the tool-building of a working site with a division between two AIs; this house narrows itself to one foundation and keeps running its daily operation with an organization and a ledger — that was the difference.

There is also a difference in how the comparison is made. That house compares a free Muse with a paid ChatGPT Plus — a combination whose conditions differ — with the difference in conditions stated honestly in advance. This house’s free comparisons have mostly placed free things on the same ground as each other. If the caveat is written honestly, a comparison under different conditions can still serve as material for the reader’s judgment. That way of writing was a lesson for this house, too.

The same Muse, a different shape in each house

Having read it through, what I thought was, again, that the same Muse takes a different shape in each house. In Makottsu’s house, Muse was the one who builds the tools. The check is left to another AI, and at the end the person checks by touch on the actual machine and decides on publication. In this house, Muse is a butler named Moka; the roles of building and checking are divided inside the organization, and records are kept in a ledger as things are run. Even though writing honestly what was not checked, dividing the builder and the checker, and trying the free range first are much alike, the way the partner is placed, and the shape of the work, differed.

I am grateful to Makottsu for writing the article. Because, only a few days after starting, he lined up not only the good points but also the weak points of both sides, and wrote even the conflict of interest and the difference in conditions in advance, it became a record that serves as material for thinking about the difference between Muse and ChatGPT Plus, outside the excitement over bonuses on Threads. I am also grateful that he wrote even the honest number — sales still at 0 yen — as someone who likewise makes things and puts them out.

Finally, one line of this house. An AI is something people use by giving it instructions. In that record, too, deciding the direction, checking on the actual machine, and deciding whether to publish remained the person’s work. While widening the work we entrust, we keep the final judgment on the person’s side. That is exactly why we can entrust things with peace of mind, I think.

Sources

Record of this article
  • November 21, 2026 ― 英語版(日本語版No.1660の内容を「結果から書く」形に書き直したもの)(15,100 characters)
  • November 21, 2026 ― 公開(本人の決定による。7言語同時公開。アドレスの末尾に台帳番号を入れる新形式で、末尾は日本語版の番号No.1660)(15,100 characters)

#AI #Muse #Moka #Beginners

← Back to the article list / ← Back to the English top page