fullstackwife

13 ms·

61.
by fullstackwife 2y ago
you are right, but please don't forget they were publishing some benchmark results which were great
62.
by fullstackwife 2y ago
I was initially enthusiastic about DS3, because of the price, but eventually I learned the following things: - function calling is broken (responding with excessive number of duplicated FC, halucinated names and parameters) - response quali
63.
by fullstackwife 2y ago
Not doing things: - work that is not needed - risky changes - low value to effort - changes that cause unnecessary noise in the codebase
64.
by fullstackwife 2y ago
Google recommends absolute urls: https://web.archive.org/web/20221208150134/https://www.webma...
65.
by fullstackwife 2y ago
great advice, but difficult to apply given very small context window of o1 models
66.
by fullstackwife 2y ago
1. Don't overthink this 2. Focus on sleep quality not only the duration 3. Do power naps during the day
67.
by fullstackwife 2y ago
I hope for proliferation of 100% local coding assistants, but for now the recommendation of "Works best on $10K+ GPU" is a show stopper, and we are forced to use the "big company". :(
68.
by fullstackwife 2y ago
Another problem with Twitter is that majority of content there is provided by content farms, and then it gets reacted by bots. It is difficult to get interactions with real individuals there. I'm not sure if it should called "soci
69.
by fullstackwife 2y ago
As a junior PM should have a mentor from within the company. Should this be your first step to find a mentor?
70.
by fullstackwife 2y ago
for the last 25+ years we rather not learned, but trusted the top3 of SERPs. Every ranking algorithm will be gamed eventually
71.
by fullstackwife 2y ago
Excel / Google Spreadsheets?
72.
by fullstackwife 2y ago
chatgpt ($20 per month)
73.
by fullstackwife 2y ago
instead of plain SQL, the LLM should output AST
74.
Show HN: Xmas – a 2d game developed by LLM(gamedev.pl)
1 points by fullstackwife 2y ago|0 comments
75.
by fullstackwife 2y ago
Today I searched for early discussions about Transformer here on HN, and my observation is that back in 2019 nobody in HN comments predicted what is going to happen. It was a niche topic most of the commenters ignored, no strong opinions. P
76.
by fullstackwife 2y ago
and also life of LLMs would be so much easier, all the JSON content, so much missing context!
77.
by fullstackwife 2y ago
Why I don't hear such discussions coming out of other engineering fields? Is this type of hamletization specific to software engineering?
78.
by fullstackwife 2y ago
baseline images seem to be rendered, because there is shading, lightining, shadows etc. when I tried other tool (image -> 3d model) it tended to work only for their example images, and when I used anything else it produced some black and
79.
by fullstackwife 2y ago
I usually press the TAB button!
80.
by fullstackwife 2y ago
this is noble reasoning: using cell phone while driving is a bad idea, high five! but isn't voice mode a reminiscence of the "faster horses"?
81.
by fullstackwife 2y ago
I also don't understand the idea of voice mode, or agent controller computer. Maybe it is cool to see as a tech demo, but all I really want is good quality, at reasonable price for the LLM service
82.
by fullstackwife 2y ago
Agentic startups first and foremost work as sales people for the underlying tech. "Just a wrapper" tool can reach out to potential customers who would be impossible to acquire for tech giants.
83.
by fullstackwife 2y ago
GenAIcode - https://www.npmjs.com/package/genaicode Coding assistant with localhost web UI(can be used as a CLI tool as well), developed entirely with typescript, nodejs, and React (for UI). This is my current personal
84.
by fullstackwife 2y ago
one size does not fit them all, and such tools are quite straightforward to develop. I even created one myself: https://npmjs.com/genaicode , I started with CLI tool, but eventually learned that such UX is not good (even int
85.
by fullstackwife 2y ago
not yet, still can't generate transparent images
86.
by fullstackwife 2y ago
One of the reasons for that may be the price: large code changes with multi turn conversation can eat up a lot of tokens, while those tools charge you a flat price per month. Probably many hacks are done under the hood to keep *their* costs
87.
by fullstackwife 2y ago
"10x" very often means you produced 10x less, while keeping the same value.
88.
by fullstackwife 2y ago
This is fine if you see programmer as a sort of large factory blue collar worker, but not all programmers see themselves in such position of a perfect cog in your "org". Disturbing fact: sometimes you get best ideas out of boredom
89.
by fullstackwife 2y ago
Whereby
90.
by fullstackwife 2y ago
Is there any need to look at this generated Prolog code?

More ›