jdm2212

4 ms·

31.
by jdm2212 22d ago
IIRC they actually don't and it materially reduced Meta's profit when Apple rolled out the advertiser ID system.
32.
by jdm2212 22d ago
I think that's definitely the right way to understand benchmark saturation, but there's a separate problem where the benchmarks are just not representative of real workflows even when they don't seem to be saturated.
33.
by jdm2212 23d ago
And Fable 5.x makes Opus 5 look pretty dim, despite benchmarks suggesting they're comparable. The benchmarks really are just kinda meaningless.
34.
by jdm2212 27d ago
Yeah, if you're actually trying to commit genocide, yeah, don't bother with facial recognition. But in a first strike you'd want to make sure that you actually kill the entire military command and control apparatus, and just
35.
by jdm2212 27d ago
You don't need mis-aligned AI to program a drone to fly to a specific address, loiter, and then dive bomb the first person whose face matches a predetermined photo. This was doable in principle with tech from four years ago. What'
36.
by jdm2212 27d ago
Israel and Iran pulled off rudimentary versions of this attack. There are so many shipping containers going into and out of every country, with typically zero inspections of any kind, that it is not actually all that hard to get thousands o
37.
by jdm2212 27d ago
Yes, we should be less worried about "alignment" of AI with the person operating the AI, and much more worried about various flavors of cheap, unmanned systems with really rudimentary (non-frontier) AI. The two compete directly wi
38.
by jdm2212 27d ago
The LW / rationalist / MIRI / safety crowd are in fact doomers who went off the deep end. They're fixated on AI itself as the risk ("alignment!!!1!1!!"), as opposed to what humans with these tools will do. We&#
39.
by jdm2212 27d ago
It's not really regulation-driven. It's just what happens when the people buying software and the people using it are different. This is the standard problem of enterprise software is marketed to the IT department (which cares abo
40.
by jdm2212 29d ago
Google search only dies if AI wins, in which case there's no shortage of demand for tokens. NVIDIA's revenue just doubled year-over-year; even if token demand cools off and they slow to say 25% growth, they'll be fine. OpenAI
41.
by jdm2212 29d ago
The original claim was "Their debt load and infrastructure costs compared to their relatively meager earnings..." which is just obviously not applicable to Google and NVIDIA, which both crank out more than $100 billion per year in
42.
by jdm2212 29d ago
A bubble is a debt-fueled run-up in asset prices driven by over-optimistic projections of future fundamental value. The housing bubble was people borrowing money (debt-fueled) to buy ever more expensive houses (rising asset prices) on the a
43.
by jdm2212 29d ago
Who is "they"? NVIDIA? Google? Anthropic?
44.
by jdm2212 29d ago
What specifically is the bubble you're talking about? Like which asset is seeing a runaway increase in valuation detached from fundamental value?
45.
by jdm2212 29d ago
Him and literally everyone else with investable money.
46.
by jdm2212 29d ago
Why would this cause a market crash? A supply crunch in AI will push prices up.
47.
by jdm2212 1mo ago
This is a pretty unimpeachable causal identification strategy finding Facebook's rollout caused worse mental health issues https://www.aeaweb.org/articles?id=10.1257/aer.20211218 It's hard to do similarly hig
48.
by jdm2212 1mo ago
Automation has been making food steadily cheaper for 300+ years, and continues to do so. Sometimes the price of oil goes up so much that fuel and fertilizer price increases negate the wins from automation, but automation is still generating
49.
by jdm2212 1mo ago
Pretty much, but ideally it should be transparent to your app developers. App developer writes `rpc.doThing(...)` on one side, and an implementation on the other, and the infrastructure -- the RPC framework or the service mesh or whatever -
50.
by jdm2212 1mo ago
Doesn't do you any good if the outage is in the circuit breaking layer, which it was for GitHub (this started as a load balancer outage).
51.
by jdm2212 1mo ago
When you have an outage, you should not retry at all. Exponential backoff is exactly how you get cascading outages. If service A fails a request to service B and decides to exponentially back off, now service A is holding open an end user r
52.
by jdm2212 1mo ago
The exciting thing about AI is precisely that it'll let software move beyond Yet Another B2B SaaS and into doing useful things in the real world. I regularly ride in driverless cars! That was the stuff of science fiction when I was a k
53.
by jdm2212 1mo ago
The right answer is for the RPC framework to accurately communicate "try again on another node" vs "don't try again, just hard fail". When one end user request fans out to hundreds of backend requests (typical for m
54.
by jdm2212 1mo ago
Try asking Opus or Fable that question. It'll give you a good answer on why microservice architectures work the way they do in order to keep user-facing latency acceptable and minimize downtime. It's a complicated enough topic tha
55.
by jdm2212 1mo ago
Exponential backoff is the wrong answer in a highly available system in the typical case where (a) failure is expected and (b) you have nodes you are supposed to fail over to.
56.
by jdm2212 1mo ago
A common pattern in highly available services is that sometimes you should retry immediately (because the node you hit is rolling/broken/overloaded, but the others aren't) and other times you should back off aggressively (bec
57.
by jdm2212 1mo ago
If your LB is down, you're just kind of screwed. You can't enqueue things if requests aren't getting through at all. Same deal with authn/authz issues, which they also had. If you can't answer the question "is
58.
by jdm2212 1mo ago
How would any of what you're saying help with this? > The immediate cause of the failure was network saturation on load balancers in Central US due to a new peak in traffic. Originally this was caused by an Istio sidecar pod reachin
59.
by jdm2212 1mo ago
The trend line does not tell you what will actually happen at scale, even if you think you're perfectly prepared for the next 10% or 20% growth. As Mike Tyson put it, "everyone has a plan until they get punched in the face".
60.
by jdm2212 1mo ago
This stuff is good! This is what a booming economy looks like. There are people out there competing with you for resources because they have cool ideas they want to implement.

More ›