Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
aiunboxed
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
9 ms
·
61.
▲
by
aiunboxed
3y ago
Because the reality is that there are very less successful adoptions of GPT on business applications.
62.
▲
by
aiunboxed
3y ago
Does this work in the open ai playground as well ?
63.
▲
by
aiunboxed
3y ago
How about mentoring junior devs in your org ? How has the experience been there ?
64.
▲
by
aiunboxed
3y ago
Open AI has a batch api for generating embeddings, rest haven't found much libraries built on top of open ai which are production ready
65.
▲
by
aiunboxed
3y ago
Yup, i was there for some time, I remember there was a accelerator on King's street. But yes pretty much occupied by big companies.
66.
▲
by
aiunboxed
3y ago
Heard the startup sync is good in waterloo, how true is it ?
67.
▲
by
aiunboxed
3y ago
Amazing, nice curation. Had a look at the other blog posts, great work.
68.
▲
by
aiunboxed
3y ago
+1
69.
▲
by
aiunboxed
3y ago
Nice, i was reading about this couple of days ago and belong to place near this city. RIP
70.
▲
by
aiunboxed
3y ago
Open AI has released this as a feature, is this news ? what am i missing ?
71.
▲
by
aiunboxed
3y ago
I am an existing aurora customer and for me also aurora is already contributing to 33% of the cost, while the team is working on optimising pgsql queries, this is a great hope. I hope xmas come early for us as well :)
72.
▲
by
aiunboxed
3y ago
Best way to focus is to not use slack.
73.
▲
by
aiunboxed
3y ago
Thanks for the detailed answer.
74.
▲
by
aiunboxed
3y ago
Which theme are you using on Hugo?
75.
▲
Ask HN: What is your blogging stack?
9 points
by
aiunboxed
3y ago
|
23 comments
76.
▲
by
aiunboxed
3y ago
I guess what you mentioned is the story of most of the people who have tech money :)
77.
▲
by
aiunboxed
3y ago
Nice, can you also please comment on the resources required to run this and the speed of the output?
78.
▲
by
aiunboxed
3y ago
Slack is really the worst, it is almost like a full blown social media. They are trying to make it less spammy but I think it is too late now.
79.
▲
by
aiunboxed
3y ago
Yes that is the problem, in half of the big tech companies GPT is blocked, and if it is blocked at your work, there is less chance that you will actually end up understanding the value it can bring.
80.
▲
by
aiunboxed
3y ago
I think there are broadly two types of people, one who understand LLMs and realise that it is actually going to change the world, others who think it is just a hype like crypto / NFTs/ meta verse. I have friends working in big tec
81.
▲
by
aiunboxed
3y ago
Very well written.
82.
▲
by
aiunboxed
3y ago
At this current moment what type of companies are preferring to build their own LLMs?
83.
▲
by
aiunboxed
3y ago
I think a company like Google would be having a lot of restrictions around where to train models and where to not. And they already have access to the whole internet, Gmail conversations would be one tiny part of it. Also wonder if they got
84.
▲
by
aiunboxed
3y ago
Agree, I doubt it would be trained. If something personal - like Gmail/ drive goes out in any of bard responses, it will be the end of bard.
85.
▲
by
aiunboxed
3y ago
I think it is not as simple as that. There is cost of living and demand supply levers as well. There is a huge salary gap between usa and Canada as well.
86.
▲
Ask HN: Has Open AI lost the battle in image generation?
2 points
by
aiunboxed
3y ago
|
2 comments
87.
▲
by
aiunboxed
3y ago
The riches are in the niches :)
88.
▲
by
aiunboxed
3y ago
claude is under limited preview right ?
89.
▲
by
aiunboxed
3y ago
Yes that is the truth the open AI API is indeed very slow and inconsistent. Have heard people say that Azure open AI integration is a little better ? Claude is also under limited preview right now. Also waiting for some reliable API which c
90.
▲
by
aiunboxed
3y ago
Thanks Alex :) Looking forward to the release
More ›