Google-ийн төслийн менежер асан Такер Брайант хиймэл оюун ухааны технологийн эрсдэл, тэр дундаа хүмүүс шүүмжлэлгүйгээр технологид найдах хандлагын эсрэг уг арт төслийг бүтээжээ.
Олон нийтийн ярианы үгээр бол хиймэл оюун ухаан буюу “AI” нь энэ тохиолдолд “дундаж хувь хүн” (average individual) гэсэн утгыг илэрхийлж байна. Гучин хоёр настай уран бүтээлч Такер Брайант уг платформоо дөрөвдүгээр сард эхлүүлсэн бөгөөд Сан-Франциско хотод сурталчилгааны самбар байршуулснаар олны анхаарлыг татаж, наймдугаар сарын 6 гэхэд 30,000 гаруй хандалт хүлээн авчээ. Хэрэглэгчид өдөр тутмын хоолны сонголтоос эхлээд хувийн чанартай олон төрлийн асуулт илгээж байна.
Зохиогч уг платформоо зөвхөн хөгжилтэй туршилт бус, харин хүмүүс өдөр тутмын энгийн шийдвэр гаргахдаа ч хиймэл оюун ухаанд хэт найдаж, шүүмжлэлт сэтгэлгээгээ алдаж буй “танин мэдэхүйн буулт” (cognitive surrender)-ын эсрэг үзэл санааг илэрхийлсэн уран бүтээл гэж үзэж байна. Хэдийгээр тэрбээр хиймэл оюун ухааныг эсэргүүцдэггүй бөгөөд Lovable хэрэгслийг ашиглан вэбсайтаа бүтээсэн ч технологийн эрсдэлийг бодитойгоор харахыг зорьжээ.
Хэрэглэгчид эхэндээ үүнийг хошигнол хэлбэрээр хүлээж авч байсан ч бодит яриа өрнүүлэх боломжтойг ойлгосноор илүү чин сэтгэлийн, хувийн шинжтэй асуултуудыг илгээх болсон байна. Тэрбээр өөрийгөө чадварлаг зураач эсвэл мэргэжилтэн биш гэдгээ тодотгож, хүмүүст мэргэжлийн зөвлөгөө өгөхөөс илүүтэйгээр зүгээр л санаа бодлоо солилцох хүний оролцоог мэдрүүлэх нь энэхүү төслийн гол үнэ цэн юм гэж онцолжээ.
Дэлгэрэнгүйг эх сурвалжаас харах
↓Эх сурвалжийг нээх ↓
What if, instead of typing the next silly question you have into Claude or ChatGPT, you asked a random guy instead?
That’s the idea behind ChatTJB. It’s the creation of 32-year-old Tucker Bryant, a public speaker, artist, and former project manager at Google. A San Francisco billboard describes ChatTJB as “the leading chat interface powered by AI.” In this case, though, AI stands for “average individual.”
Bryant didn’t train an LLM to respond based on how he writes. He literally sits down and answers the prompts himself. If you request an image of a skunk wearing sunglasses, he’ll scribble one up. If you ask how to fry an egg, he’ll respond with a suggestion. Maybe. (WIRED asked ChatTJB how to become a morning person, and at the time of writing, we haven’t received a response. The website notes that “thousands of people are writing in right now.”)
Bryant tells WIRED he launched ChatTJB back in April, but it didn’t take off until July 27, when his $6,000 billboard went up in San Francisco. News coverage soon followed. As of August 6, he says, ChatTJB has received over 30,000 queries. At its peak, he says, he was receiving about 5,000 prompts per hour.
Bryant says that he “first and foremost” considers ChatTJB to be an art project, not just a funny bit or a social experiment.
“It was sort of an opportunity for me to provide a little bit of commentary on a strange moment that I feel like we’re in,” Bryant says. He doesn’t consider himself anti-AI—in fact, he vibe-coded the ChatTJB website using Lovable—but he’s clear-eyed about the risks of the technology.
He’s concerned about the risk of “cognitive surrender,” which refers to accepting AI outputs with increasingly little critical thinking or scrutiny. He says the concept made him reflect on his own back-and-forths with LLMs, like the time he asked if he should wear short or long sleeves on a 65-degree day in San Francisco.
“And then I’m like, dude, I’ve been living in this city for whatever, like seven years,” Bryant says. “Like, I know how to handle the weather. And yet, it felt like an easier thing for me to defer to a LLM for advice on that.”
“It felt like something that I wanted to speak on,” he adds, “because I was like, what am I doing here?”
Over the past couple of days, he has drawn a frog eating spaghetti, San Francisco mayor David Lurie as Godzilla, and two black cats eating grass on a table, among other things. The written requests, Bryant says, have been diverse.
“People were asking things like, ‘Oh, it’s my birthday in two weeks, and I don’t really want to have a party—what do you think I should do?’” Bryant says. Other people ask what they should have for dinner or how they can use the ingredients in their fridge.
“A lot of people initially were interacting with this thing as a joke or as a funny thing,” Bryant says. “And then I think when they realized that they could chat and get actual conversation, some of what people would share would be more personal, sincere, genuine, et cetera.”
Bryant is quick to note that he’s not a good artist, and his advice is no better than any other person’s. (On YouTube, he described it as “the worst LLM of all time.”) But he thinks people understand that.
“The way I’m interpreting it is that the value for them is just getting a human to bounce those questions off of, as opposed to getting super expert advice,” Bryant says. “It’s very tender to see.”
