Frederic Marx
@fmarx.com
Enthusiasm enthusiast. Frontend developer at Ableton. Writes about code, design, music, and general nerdery.
Reposted by Frederic Marx
Another view of ITC Zapf Dingbats in U&lc, vol. 5, no. 2, 1978. Layout probably by Herb Lubalin.
archive.org/details/ulc-...
archive.org/details/ulc-...
October 25, 2025 at 8:19 AM
Another view of ITC Zapf Dingbats in U&lc, vol. 5, no. 2, 1978. Layout probably by Herb Lubalin.
archive.org/details/ulc-...
archive.org/details/ulc-...
I see an opportunity cost here.
Whenever we agree that an LLM produces results that are “good enough” (whatever that means), we give up on the possibility of better things.
Conveying the meaning of an image in text is an artful craft, but if we say, “just use AI”, why should anyone invest in it?
Whenever we agree that an LLM produces results that are “good enough” (whatever that means), we give up on the possibility of better things.
Conveying the meaning of an image in text is an artful craft, but if we say, “just use AI”, why should anyone invest in it?
October 13, 2025 at 1:42 PM
I see an opportunity cost here.
Whenever we agree that an LLM produces results that are “good enough” (whatever that means), we give up on the possibility of better things.
Conveying the meaning of an image in text is an artful craft, but if we say, “just use AI”, why should anyone invest in it?
Whenever we agree that an LLM produces results that are “good enough” (whatever that means), we give up on the possibility of better things.
Conveying the meaning of an image in text is an artful craft, but if we say, “just use AI”, why should anyone invest in it?
Reposted by Frederic Marx
