popinman322

11 ms·

61.
by popinman322 4y ago
IIUC -> If I understand correctly
62.
by popinman322 4y ago
IIUC: Spotify seems to use something similar to collaborative filtering. People rate songs based on more than similarity, which is what this model seems to provide.
63.
by popinman322 4y ago
Have you tried Gnoosic[0] yet? It's a more standard recommender system, I think. Some of the recommendations are really on point. [0]: https://www.gnoosic.com/
64.
by popinman322 4y ago
As a mid-level dev, I remember putting $40/month in the survey OpenAI sent out. Granted, I kind of also expected API access for that price. It's a very interesting multi-tool; being able to write first drafts for docs, summarize n
65.
by popinman322 4y ago
Properly creative work still isn't automated. The current generation of systems can only generate content within the codomain (semantic space) they were trained on; they're interpolating and need more work to extrapolate. Most cre
66.
by popinman322 4y ago
This reminds me of a discussion I had with a friend while playing Factorio together. Traditionally you'd put a list of recipe ratios into a spreadsheet and then calculate what you need. But there's a mod called Helm that can handl
67.
by popinman322 4y ago
What kind of basic math are we talking about here? Arithmetic? Calculus? Group theory? What makes those "basic maths" a good benchmark for whether we're fucked? Freeing up head space for more important things is valuable. Not
68.
by popinman322 4y ago
> Maybe by parsing through all related meeting transcripts as well? Those and design docs. It can also reason over code somewhat; it may be possible for the model to intuit design constraints from code.
69.
by popinman322 4y ago
Snippets of your local code are sent to the API as part of the Copilot request context.
70.
by popinman322 4y ago
PCA is a foundational dimension reduction technique, and kNN can be used in conjunction with embeddings. k-means is still great when you have prior/domain knowledge about the number of groups.
71.
by popinman322 4y ago
I know there's some active work on this (using LLMs, not traditional methods), not on the binary side but on the source analysis side. See https://grit.io/ , which tries to detect bugs (and maybe vulnerabilities?) and a
72.
by popinman322 4y ago
I've been waiting to see something like this. There's certainly room to fine-tune an LLM for this task; in that vein, I wonder whether Ghidra's pcode would produce better results? It's a bit better suited to this task in
73.
by popinman322 4y ago
You can already coax ChatGPT into interacting with external systems today; I set up a prompt where the model pretended to be a factory system on a communication bus. It could access its "inventory" by posting a prefixed message to
74.
by popinman322 4y ago
IIRC, ChatGPT is based on GPT3.5 (likely an even larger model) rather than GPT3. It's also been refined a bit using reinforcement learning. I've noticed that when I ask ChatGPT to determine the type of a variable in a given code b
75.
by popinman322 4y ago
Seems like speed wouldn't be that much of a problem when the satellites all move in a highly predictable manner.
76.
by popinman322 4y ago
> while dating apps feel loser-ish for people who can't socialize but are desperate. Apps allow you be to visible to other people who are also open to getting together. Not all of the "infinite supply of women" in your are
77.
by popinman322 4y ago
It could also be a useful proxy for density/interestingness of the material (when adjusted for your typical reading speed). Fed into a recommender system it might produce better outputs and more purchases. It could also be sold to publ
78.
by popinman322 4y ago
Ghidra still mostly relies on community plugins for obscure platform support. It's ~OK but less than ideal.
79.
by popinman322 4y ago
> "went $FOO miles in $BAR conditions without needing any intervention" But who decides whether intervention was necessary? This is still too vague.
80.
by popinman322 4y ago
At least a year ago in Azure that wasn't true; everyone had access to nearly every internal service's code (+the windows kernel). Though there were some exceptions (the Teams team didn't want to share their source at all for
81.
by popinman322 4y ago
They're likely referring to "storing your config data in a repo" as "dotfiles" here. It makes sense in that light: where backing up your config data makes life easier for you when you set up a new machine, this make
82.
by popinman322 4y ago
I agree that the opportunity cost of dropping 100B on this is very high, but what if the new data enables tech that can't otherwise be achieved? We could definitely postpone this work and invest in more impactful work in the present. I
83.
by popinman322 4y ago
The LHC's development produced a number of new technologies (e.g. radiation resistant microcontrollers, and gas electron multipliers). Could the next generation produce more, valuable tech? I also don't have the expertise to tell
84.
by popinman322 5y ago
Retinoids: your doctor Everything else: The Ordinary
85.
by popinman322 5y ago
It's true enough, but it never really felt like teams solicited feedback on their APIs or portal UX. Azure really only made this jump to using all public products internally recently; even the CI systems used by internal teams were pro
86.
by popinman322 5y ago
I imagine I'll be editing this a bit, so I apologize if there are obvious typos left from any changes I make while I'm thinking. Sorry for the mini-essay. :) Also, these points are not to be taken separately. They're part of
87.
by popinman322 5y ago
Have you contacted the support team yet? Escalation likely requires an open ticket.
88.
by popinman322 5y ago
Looks like DC power is more efficient for long haul cables. https://engineering.stackexchange.com/a/19759
89.
by popinman322 5y ago
Can confirm, my Presonus Audiobox works well enough and passes through just fine to the windows VM I use for AV stuff. Still needed another amp though.
90.
by popinman322 5y ago
I'm generalizing my own experience here, but I imagine it typically takes time for people to reach the conclusion that they want to leave. What if they enjoy some of those "un-fun" tasks? Or maybe they take pride in their wor

More ›