causal

7 ms·

91.
by causal 3mo ago
Yeah I sometimes see people on here getting defensive when you call out AI slop, saying maybe it's just a human who writes like Claude, and I really don't care- slop is slop.
92.
by causal 3mo ago
That battle tested bit is so true, always was. It’s hard for engineers to admit their hard work is worthless until they have let other people use and shape it.
93.
by causal 3mo ago
I think it was always hyperbolic to call rewrites the single greatest mistake (I can think of worse) but I think most of the wisdom still holds. And AI generated rewrites are arguably riskier because now NOBODY is familiar with the massive
94.
by causal 3mo ago
If OpenAI is this shady culturally then all kinds of dark secrets could come out. They just released 5.6 to great fanfare with benchmarks showing even their weakest model (Luna) supposedly smoking Fable. Having used it a couple hours I&#x27
95.
by causal 3mo ago
I feel like the converse question is more relevant: what projects have open source agents been good enough to implement on their own?
96.
by causal 3mo ago
Distillation is a specific named technique from a pretty famous paper "Distilling the Knowledge in a Neural Network," by Hinton and others. What makes it different from just training on any data is that you train the student model
97.
by causal 3mo ago
So you can also destroy trust other ways. What’s your point?
98.
by causal 3mo ago
Still debating whether these are actually information dense or not. Complexity gives the appearance of density, but these sentences always come in whole pages at a time while often failing to deliver the needed information.
99.
by causal 3mo ago
I wish people would stop using Anthropics incorrect use of the term distill. They don’t share logits so you can’t distill. You can generate training data, which doesn’t sound nearly so scary.
100.
by causal 3mo ago
Way easier said than done, and hiding that behavior isn’t trivial, and huge waste of compute budget if it’s found and never used. Also not difficult to run in contained environments where it doesn’t have access to Internet to begin with. No
101.
by causal 3mo ago
Yeah I recently downgraded my subscription. The paternalism is out of control. Secret weights, secret guardrails, secret stenography, secret dumbing-down of triggers, secret token allowances.
102.
by causal 4mo ago
Yeah though to be fair that's a settings option.
103.
by causal 4mo ago
Depends on leadership culture. In toxic (aka “competitive”) environments managers are insecure and fear their own staff as potential competition.
104.
by causal 4mo ago
For anyone wondering what the actual purported security weaknesses are in this article (I used the slop machine to reduce the slop): - Cloud backups — by default, backups to iCloud/Google Drive contain plaintext messages, and E2EE back
105.
by causal 4mo ago
I think it's important to multiply "likelihood of the human making a mistake" by "how many times the human must avoid a mistake". If your hobby presents 100 avoidable-but-life-threatening mistakes an hour, it's
106.
by causal 4mo ago
Too many people read "human error" as "human preventable" rather than "a thing you will also do because you are human"
107.
by causal 4mo ago
So if it works: Awesome. The spa approach is a little weird. FDA workaround?
108.
by causal 4mo ago
Absolutely. Imagine forking VSCode and then selling it for $60 BILLION in a few years. And right as you lose your competitive advantage to foundation providers. Amazing.
109.
by causal 4mo ago
Okay then makes me wonder if this recent trend is just one particularly large manufacturer ramping up production? Tesla?
110.
by causal 4mo ago
But it's just not a high quality essay. This writing would have few if any upvotes coming from a random Substack blog. I still really like some PG essays, but PG used to be a much better writer.
111.
by causal 4mo ago
> I can't speak for his other essays but this article was not great. Indeed this wouldn't even be front page material if PG's initials were not on it.
112.
by causal 4mo ago
I mean there's also just sheer luck. Getting billions is probably more about luck than hard work OR corruption, even if those two things can increase your chances. But luck is also not "earning" in most senses of the word.
113.
by causal 4mo ago
I thought he might be onto something interesting when he began with the growth-rate example, since that's most-obviously tied to valuation. But he skipped over valuation entirely and decided to lecture a bunch of Oxford grads about mid
114.
by causal 4mo ago
I think maybe a simpler explanation is that tech has been such a story of purportedly humble people becoming wildly successful. Classic rags to riches. Makes it easy to think of nerds as one of the common people, even the rich ones.
115.
by causal 4mo ago
Sure but you can totally contrive scenarios to give the appearance of what you described without really doing anything notable. What matters is scale. Did it deploy a novel zero-day exploit to overcome a problem? That's alarming. Did i
116.
by causal 4mo ago
One thing I find kind of annoying is how Anthropic goes for these "vast and alien" names like Fable and Mythos, but then deliberately trains the model's personality to act like a cool high school teacher that feels totally fa
117.
by causal 4mo ago
This depicts a kind of "dark forest of AI agents resorting to kill or be killed" narrative but it sounds more to me like an agent just earnestly problem-solving why its processes are being killed without real awareness of what was
118.
by causal 4mo ago
Sure but finding their shortcomings and patching them with skills takes real trial and error. They are incapable of identifying their own shortcomings for you.
119.
by causal 4mo ago
I can't even get Claude or GPT-5 to consistently produce good flows for common use cases, much less domain-specific shit. They have deep vocabulary though, which makes them sound better informed than they are. They are very good at wri
120.
by causal 4mo ago
> to train in a role that’s insulated from AI Would love to know more about that role

More ›