iknownthing

7 ms·

31.
by iknownthing 2y ago
Can Claude create and run a CI/CD pipeline now from a prompt?
32.
by iknownthing 2y ago
I did something very similar to this, except I would say I did learn something. It felt like studying by looking up the answers - not the best way to study but if you look up the answers to enough questions you are bound to learn something
33.
by iknownthing 2y ago
That video looks like what I need, I'll give it a try, thanks
34.
by iknownthing 2y ago
I wish they were open about their ghosting.
35.
Ask HN: LLM Router Recommendations
3 points by iknownthing 2y ago|4 comments
36.
by iknownthing 2y ago
About a year or so ago I interviewed with Sourcegraph for about 6 months then they ghosted me and the role I was up for disappeared. Pretty sure no one got hired for it.
37.
by iknownthing 2y ago
Is cursor the new co-pilot now?
38.
by iknownthing 2y ago
I building a web app around a tool I originally built for myself for interview prep. It basically lets you generate interview questions then practice them in a leetcode-like interface that includes an AI tutor chat ui. The tutor is aware
39.
by iknownthing 2y ago
How about just "misprediction"?
40.
by iknownthing 2y ago
Something I'm not sure I understand. It seems there are a set of expectations that you are evaluating against. Are these expectations also written in the prompt?
41.
by iknownthing 2y ago
Completely agree. The vast majority of the failed projects fail at the planning stage. To put it simply, if the cost of a misprediction is high then it is usually not going to work because all models make mispredictions. This is somethin
42.
by iknownthing 2y ago
To be clear, I had many interviews with them over that 6 months (10+ I think).
43.
by iknownthing 2y ago
I interviewed with them once, they strung me along for about 6 months then ghosted me.
44.
by iknownthing 2y ago
bait & switch jobs should also be fraud
45.
by iknownthing 2y ago
Is it possible to embed the stempad editor in something like a react app?
46.
by iknownthing 2y ago
and even if there is another breakthrough all of these companies will implement it more or less simultaneously and they will remain in a dead heat
47.
by iknownthing 2y ago
I wish we had a definition for machine learning engineer
48.
by iknownthing 2y ago
for some reason I absolutely love the name
49.
by iknownthing 2y ago
I'm curious how you run the python code in the browser
50.
by iknownthing 2y ago
I saw the k-means one a couple times
51.
by iknownthing 2y ago
I've definitely had ML questions involving coding e.g. implement k-means
52.
by iknownthing 2y ago
The difference is the frontend role would require a real understanding of React, but these "AI dev" roles don't require a real understanding of AI.
53.
by iknownthing 2y ago
I'm always skeptical of these roles. Very often they do not involve anything AI related which is why a formal AI background is not required.
54.
by iknownthing 2y ago
at what point does a goldrush become a shovelrush?
55.
by iknownthing 2y ago
So AI infrastructure startups are tarpits and so are the wrappers? Is it just tarpits all the way down?
56.
by iknownthing 2y ago
Agreed. The LLMs have effectively been democratized. As much as people deride "LLM wrappers", the quality of the wrapper and how creatively they use the LLM api is the differentiator.
57.
by iknownthing 2y ago
> tired of the infantilization of the industry What does this mean?
58.
by iknownthing 2y ago
Looks like they've parlayed it into some kind of business https://www.langchain.com/
59.
by iknownthing 2y ago
I tried LangChain a while ago for a RAG project. I liked how I could just plug into different vector stores to try them out. But I didn't understand the need for the abstractions around the API calls. It's not that hard to just
60.
by iknownthing 2y ago
Seems like you need to have been working at a place like Google too

More ›