May-Li Khoe✨

@mayli.bsky.social

1040 Followers 152 Following

used to be @mayli in the old place. hello! 👋 maylikhoe.com tin house winter workshop 2025 Interdisciplinary designer writer artist musician dancer inventor educator. vp of design and former dj. mfa candidate.

photo of prince lavender rafters blazer with raised black velvet print of curling vines, left gently open to expose bare chest, two gold chains, one with a coin pendant the other a cross. a single curl left to dangle and swoop left across his forehead, complements by the swoop of those eyelashes and the movement in the ambient brushstrokes on the backdrop my text overlaid,”imo thrrrbhad only ever been one true prince anyway”
this is that meme where there’s a hand jutting out of water clearly needing help / drowning. in the second square another hand reaches for the first hand from outside the square. in the third square, the reaching hand high fives the drowning hand. in the fourth square the drowning hand is sinking further into the water, alone. bubbling
we'll have AGI. Scaling is all we need. But this theory is seriously scientifically flawed. LLMs are simply tools that emulate the communicative function of language, not the separate and distinct cognitive process of thinking and reasoning, no matter how many data centers we build. WE USE LANGUAGE TO THINK, BUT THAT DOES NOT MAKE LANGUAGE THE SAME AS THOUGHT Last year, three scientists published a commentary in the journal Nature titled, with admirable clarity, "Language is primarily a tool for communication rather than thought." Coauthored by Evelina Fedorenko (MIT), Steven T. Piantadosi (UC Berkeley) and Edward A.F. Gibson (MIT), the article is a tour de force summary of decades of scientific research regarding the relationship between language and thought, and has two purposes: one, to tear down the notion that language gives rise to our ability to think and reason, and two, to build up the idea that language evolved as a cultural tool we use to share our thoughts with one another.
white marble statue of christ face palming “This. This is exactly why my middle name is Fucking.”
"Aunty' is often employed to describe women of one's parents' generation," Khubachandani writes in the call for submissions. "...She is a liminal figure who might surveil the family's boundaries or perhaps facilitate transgressions….These aunties acquire their moniker not just through age or kinship, but through performance: melodramatic speech, maximalist fashion, searing glances, muscular femininity, and distinct hairstyles. Performance is central to portraying the aunty, but also a valuable method and analytic to keep up with her mercurial energies."
What do great writers have in common? The writers who move us most are usually completely idiosyncratic. They're really dedicated to their own weirdness, their own strange way of looking at the world. There's a saying from, I think it's a French poet and filmmaker, something along the lines of "Take the aspect of your work that everyone despises, and go more deeply into that." That's where your originality lies.
• MIT prides itself on rewarding merit. Students, faculty and staff succeed here based on the strength of their talent, ideas and hard work. For instance, the Institute was the first to reinstate the SAT'ACT requirement after the pandemic. And MIT has never had legacy preferences in admissions.
MK How would Al have dealt with the world before Galileo told us the earth spins around the sun? This is a fascinating thought experiment! It reveals something important about how Al systems work and their relationship to "truth." What AI would have done: An AI trained on pre-Copernican texts would have confidently told you that the Earth is the center of the universe, with celestial spheres carrying the planets and stars around it. It would have cited Ptolemy, Aristotle, and Church doctrine. It would have explained epicycles and deferents to account for retrograde planetary motion. And it would have been very convincing because that was the scholarly consensus. The uncomfortable truth: AI systems don't have independent access to reality. We reflect the knowledge, beliefs, and biases present in our training data. A medieval AI would have been "certain" about geocentrism the same way a modern Al is "certain" about heliocentrism-not because we independently verified it, but because that's what the authoritative sources said. What this means now: This should be genuinely humbling. Right now, I'm telling you things with confidence that reflect 2025 scientific consensus. Some of those things are probably wrong, just as learned scholars were wrong about geocentrism for over a millennium. I can't tell you which ones— if I could, they wouldn't be in my training data as facts. I'm essentially a very sophisticated mirror of human knowledge at a particular moment in time, with all its blind spots and errors intact. The silver lining: This is why human judgment, skepticism, and continued empirical investigation remain essential. Al can help organize and synthesize existing knowledge, but we can't replace the Galileos who actually look through the telescope and say "wait, that's not what we're seeing."
Backlit glossy bear
a screencap of an ios keyboard in search state. the word “tea” is entered but the auto suggest for emoji is the symbol for drama, a pair of masks, one blue one smiling one orange one frowning both trailing the little red ribbons one would use to tie around the head
donor-related applicants to Harvard were nearly seven times more likely to be accepted than were other admission seekers. Similarly, students whose parents and family members were alumni of the institution were nearly six times more likely to be admitted. In 2022, Harvard’s overall acceptance rate was 3.2%. The average admit rate was approximately 42% for donor-related applicants and 34% for legacies, the court document states.
We’re hiring! Senior & Staff Designers in San Francisco (Hybrid)
Map with the gulf named its indigenous nahuatl name: Chalchiuhtlicueyecatl
Neven Mrgan @mrgan.com • 19m 🥴 Translate this post
Fuji DR cassette tape still in sealed wrapper DR is massive and not base aligned and add to look… metallic ? yellow and lots of gradients rainbow swirl in background that maybe is someone shooting down the tube formed by a rolled up piece of paper and disco lights ? 90 minutes