useful hands and a cheerful heart | @_torontosociety
Toronto
- Tweets 14,348
- Following 1,472
- Followers 3,822
- Likes 44,428
Pinned Tweet
I'm giving a talk next Thursday.
For anyone who came to the discussion a while back on tacit knowledge, in some ways this is a continuation of that conversation. An update on what I've been researching and thinking about in the interim.
I would love to see you there.
Join us next Thursday for a new experimental series of Toronto Society Talks.
An original thesis, rigorous questions, food, drinks and a party.
The first talk is from @_benjaminparry
Why Everyone Should Study Theology
wygo.world/society-talks-1
5
4
39
1,194
Slides are starting to come together. Excited to discuss with everyone this Thursday.
Join us next Thursday for a new experimental series of Toronto Society Talks.
An original thesis, rigorous questions, food, drinks and a party.
The first talk is from @_benjaminparry
Why Everyone Should Study Theology
wygo.world/society-talks-1
10
656
Benjamin Parry retweeted
Time to start sharing: I'm working with an unbelievably talented team to build @cypress__labs
11
22
66
11,319
Benjamin Parry retweeted
web design & brand refresh for @_torontosociety by me 🤍
13
5
130
2,620
Benjamin Parry retweeted
A new home for the Toronto Society on the internet.
Check it out at thetorontosociety.com
3
2
37
3,768
Benjamin Parry retweeted
Join us next Thursday for a new experimental series of Toronto Society Talks.
An original thesis, rigorous questions, food, drinks and a party.
The first talk is from @_benjaminparry
Why Everyone Should Study Theology
wygo.world/society-talks-1
4
16
2,375
Benjamin Parry retweeted
interpretability might be the most important scientific problem in the world right now, as I don't think we can solve alignment if we don't know what we're building. I think I have good news about it. I co-authored early work in mechanistic interpretability at OpenAI and came away pessimistic, because neural networks appeared to fold into arbitrary tangled knots. years later, I started openai's first large-scale automated interpretability project, finding about 2% of neurons legible with the primitive tools we had
with modern agentic tools and a much closer look, the picture looks totally different. the vast majority in models up to trillions of parameters are legible, and their circuits beautiful. many famous examples of tangled neurons turned out to be misunderstood. The roadblocks I expected mostly aren't there. I now think ambitious interpretability will be solved by the time we need it most, and possibly quite soon, as automated scientists keep improving. it may be a bitter lesson for interpretability: the simple thing, at absurd scale, works
already-clean.goodfire.com/
863
2,605
19,647
4,371,036
Benjamin Parry retweeted
Great conference at @rootsinprogress this weekend, as has been true in '24 and '25. What works here: 1) Everyone is on board with "high-agency, high-progress, high-optimism", but conditional on that, huge differences in field of work, political preferences, country of origin. 1/4
3
5
83
5,323
Clement is a true original
Clement Leung makes food videos as insta.noodls. He is an investigator and storyteller whose focus is uncovering dishes and cuisines worth trying.
We started by cooking Shin Ramyun on the table, and he showed us his hack: crack two eggs into the leftover broth and microwave it. Clement walked us through how he started documenting food and why.
Then we talked about Toronto, which Clement thinks might be the best food city in the world because it's the most diverse. We got into the economics of running a restaurant, how competitive it is to start one, and the food court stalls of Markham. Clement also reflected on the journey of documenting food, shared the stories he's picked up from getting closer with restaurant owners through his work, and told us what helping Mark Wiens taught him about what makes something legendary.
The Other Stuff is hosted by internetVin, filmmaker and entrepreneur and possibly the most curious man on Earth. Produced by New.
Timestamps
00:00:00 Becoming the Ramen Guy
00:04:25 Cooking Shin Ramyun
00:23:30 Gong Zai Mein and Curry Cubes
00:31:28 Stitching Through the Noodle Era
00:42:16 Best Food City in the World
00:56:29 A Map Drawn by Immigration
01:10:00 The Plaza Is a Tournament
01:22:24 No Ratings, No Free Meals
01:30:40 One More Year, Then Mary Brown's
01:45:37 Bifana, Shawarma, and Knafeh
01:57:22 Mark Wiens
02:16:59 Mark's Legendary Toronto Crawl
02:38:09 China, Ten Years Ahead
02:47:46 Eating Through China
03:01:30 Respect the Dish
2
4
610
This was great fun. Amazing to continue to watch @buildcanada go from strength to strength.
Building a brighter future for Toronto starts with showing up. 🙋
Thank you to everyone who joined us tonight at @ada_cx HQ to chat about the bold ideas that will build the Toronto of tomorrow. 💙
1
18
646
Every time I walk onto a plane sitting on the tarmac at Billy Bishop as the sun is rising I feel like a character in an Ayn Rand novel.
Let us boldly ascend to the sky. We are children of the heights. In good time we will be humbled and fall but we must reach for the heavens first.
2
35
1,232
Benjamin Parry retweeted
I’m looking for 6 Toronto families to try Torch, the instrument we’re building to make playing and singing easier.
Beginners or lapsed players. Borrow one for Nov; help us learn what makes you keep playing—or stop.
Email me: daniel@debowmusic.com
18
12
125
8,872
This will be a great session to reflect on what Toronto could be. Gosh this is a wonderful city and it could be so much more. Hope to see folks there.
Join me on Wednesday in Toronto with @_benjaminparry, Adrian Rocca (Fitzrovia), @gregbradyx
we’ll be discussing ideas to build the Toronto we know is possible
1
16
631
Fascinating paper on tacit knowledge, and mercifully only 12 pages long. Thanks @danielrock for recommending!
"Nonaka argued that ambiguity, conflict, and breakdown can force members of an organization to reexamine what they take for granted."
This made me think about how people typically characterize 'closing' the RSI loop. One interesting implication is that some of the value of human-machine interaction isn't reducible to the individual tasks being automated, but comes from the friction between different kinds of knowledge.
Specifically, while a lot of ML R&D tasks are automatable, deciding what counts as a better model will often depend on external feedback loops that unfold over time, including those that expose deficiencies in the very objectives being optimized.
arxiv.org/abs/2606.29227
8
32
244
11,864