bthornbury

6 ms·

1.
by bthornbury 17d ago
Is the tradeoff of the parallel output that we don't get arbitrary string generation? like output # of tokens is fixed ahead of time? Either way, really cool and impressive.
2.
by bthornbury 2mo ago
"Suddenly, waterfall is the thing to do." Matches my experience. Like 90% of the work is the spec. Except instead of weeks researching its like a few hours talking with an agent.
3.
by bthornbury 2mo ago
lots of good ones still
4.
by bthornbury 2mo ago
I discuss the testing approach and coverage with the model before and after, sometimes in a fresh thread that does a static analysis. interestingly my input is still pretty important
5.
by bthornbury 2mo ago
codex pro plan currently
6.
by bthornbury 2mo ago
for me, almost all of the work is specs I am no longer: - reading docs for hours and hours - typing (barely at all) - writing code - manually doing tight debug loops - using an IDE to do this I had to give up reading or even controlling the
7.
by bthornbury 4mo ago
we need some better standard long-context benchmarks. needle in a haystack is not good for this, yes it proves the model can attend to its context, but in its usual form, somewhat trivializes the query-key relationship. something like long-
8.
by bthornbury 4mo ago
the qwopus 27b model is good for grunt work style tasks, even across multiple files. Piping a bunch of things through, small factoring changes, stuff that just takes time to type out. I wouldn't rely on it for large stuff like codex th
9.
by bthornbury 4mo ago
promote yourself to PM only and use agents for authoring, verification, tests, checking the tests orchestrator -> parallel subagents with investigation, authoring, verification, benchmarking subagents and integration / final verific
10.
Show HN: Sampler Step Explorer – for understanding diffusion sampler updates(bryanthornbury.com)
2 points by bthornbury 7mo ago|0 comments
11.
An Intuitive Understanding of AI Diffusion Models(bryanthornbury.com)
2 points by bthornbury 7mo ago|1 comments
12.
by bthornbury 7mo ago
The classic papers describing diffusion are full of dense mathematical terms and equations. For many (including myself) who haven’t stretched those particular math muscles since diff eq class a decade or so ago, the paper is just an opaque
13.
by bthornbury 7mo ago
Something like a perplexity/log-likelihood measurement across a large enough number of prompts/tokens might get you the same in a statistical sense though. I expect those comparison percentages at the top are something like that.
14.
by bthornbury 7mo ago
AFAIK seed determinism can't really be relied upon between two machines, maybe not even between two different gpus.
15.
by bthornbury 7mo ago
Is modelwrap running on arbitrary clients? I'm not following the whole post, but how are you able to maintain confidence in client-owned hardware/disks following the secure model the method seems to depdend on?
16.
by bthornbury 8mo ago
Why does there seem to be such a divide in opinions on AI in coding? Meanwhile those who "get it" have been improving their productivity for literally years now.
17.
by bthornbury 8mo ago
> got a load of ticking time bomb bugs Lots and lots of tests!
18.
by bthornbury 8mo ago
Either really comprehensive tests (that you read) or read it. Usually i find you can skim most of it, but like in core sections like billing or something you gotta really review it. The models still make mistakes.
19.
by bthornbury 8mo ago
AI is getting to the game-changing point. We need more hand-written reflections on how individuals are managing to get productivity gains for real (not a vibe coded app) software engineering.
20.
by bthornbury 9mo ago
I'm not too sure about this take. The larger code rewrite issue is constantly trying to be solved, which is somehow making the problem worse. In another view, standard libraries do a pretty good job.
21.
by bthornbury 1y ago
I'm pretty sure it's called "reading the code". That said, it is difficult enough in its own right.
22.
by bthornbury 1y ago
This generalization issue in RL in specific was detailed by OpenAI in 2018 https://arxiv.org/pdf/1804.03720
23.
by bthornbury 2y ago
Recently, I've been using a local docker container to house the interpreter for all of my new python projects. For day-to-day work it is far superior than endless virtualenvs clogging up my harddrive and hunting for brew package depend
24.
by bthornbury 3y ago
Note that I had to remove two of the test cases to fit in the HN character limit: { name: "Large Input Slice", input: []any{"A", "B", "C", "D", &
25.
I have been using Mixtral everyday for coding and I think it has saved me days
3 points by bthornbury 3y ago|2 comments
26.
OpenAI cuts prices for GPT-3 by two thirds amidst growing competition(mixed-news.com)
2 points by bthornbury 4y ago|0 comments
27.
Coding Deep Q-Learning in PyTorch(youtube.com)
2 points by bthornbury 6y ago|0 comments
28.
Fixing nohup over SSH on CoreOS with systemd(aegisblade.com)
2 points by bthornbury 7y ago|0 comments
29.
by bthornbury 8y ago
Might be able to afford an apartment
30.
by bthornbury 8y ago
As an Erlang fan, who hasn't used it in production services, I'm wondering if you can let us know some specific scaling issues you encountered.

More ›