Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
adeelk93
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
7 ms
·
31.
▲
by
adeelk93
3y ago
Use the function calling API - works really well for getting structured responses
32.
▲
by
adeelk93
3y ago
Everybody knows GPT hallucinates - so occasionally it suggests code, you try it, and discover the functions or arguments it’s suggesting don’t exist. Code interpreter takes you out of that loop - GPT can find and fix its own mistakes.
33.
▲
by
adeelk93
3y ago
Two things: - your 18 year figure is inaccurate, I know people who have made it to the top of the pay scale in half that time - GS-15 step 10 is not the cap for a role like this one. this is scientific/technical expertise, which is 20%
34.
▲
by
adeelk93
3y ago
Those limits seem to be decreasing every year though. Lot of competition in the analytical SQL space. Also worth noting Delta Live Tables, another very popular streaming SQL implementation. https://www.databricks.com/product
35.
▲
by
adeelk93
3y ago
I am a bit suspicious because ChatGPT knows which version of Python it has installed, as well as which packages, without actually executing any Python. And yet, this context is not in the shared system prompt for advanced data analysis.
36.
▲
by
adeelk93
3y ago
North America’s bird population has declined by 3 billion, or 30%, in the last 50 years https://www.birds.cornell.edu/home/bring-birds-back/
37.
▲
by
adeelk93
3y ago
Why is that frustrating? I find it validating
38.
▲
by
adeelk93
3y ago
If you’re interested in doing the reverse of this (replacing pyspark with sql) - Sqlglot can do this: https://sqlglot.com/sqlglot/dataframe.html
39.
▲
by
adeelk93
3y ago
Taxis tend to cost half as much as Uber; if I see a taxi, I’ll take one to save some money
40.
▲
by
adeelk93
3y ago
Uber sold off that division years ago, and is no longer focused on self-driving https://arstechnica.com/cars/2020/12/uber-sells-self-driving...
41.
▲
by
adeelk93
3y ago
People pay more for digital media than physical media in other contexts for the convenience. I'm not going to buy a DVD on amazon to save a few dollars. It's subjective, I'm sure there's still people buying DVDs, but you
42.
▲
by
adeelk93
3y ago
I’ll endorse one - been using Airalo for a few years, works great. I prefer using that app to finding someone at the airport to buy a SIM card from.
43.
▲
by
adeelk93
3y ago
Not complying with the tax prep company’s requests is a bit suspect - we’ll see what happens, but it wouldn’t be impossible that Meta knew they had sensitive data and were taking advantage of it.
44.
▲
by
adeelk93
3y ago
In this conspiracy theory - why wouldn’t the intelligence agencies get tax returns from the IRS directly?
45.
▲
by
adeelk93
3y ago
There’s free with ads and premium without ads, what else could you want out of a business model?
46.
▲
by
adeelk93
4y ago
Couldn’t you make that argument against all SaaS?
47.
▲
by
adeelk93
4y ago
Awesome, thanks for the reply! The fallback approach sounds like it could be really great, hardening against modals and popups can be really annoying.
48.
▲
by
adeelk93
4y ago
I really like this, but I wonder if a better approach would be to take that simplified DOM and instead generate playwright or puppeteer code instead of direct DOM manipulation. That way it’s reproducible browser automation.
49.
▲
by
adeelk93
4y ago
As someone who is not familiar with these sorts of accessibility challenges, what makes copilot voice terrible?
50.
▲
by
adeelk93
4y ago
At a glance this looks like closed-source langchain - is that a fair comparison?
51.
▲
by
adeelk93
4y ago
Check out dbt - it's a great tool for organizing queries and solving such patterns
52.
▲
by
adeelk93
4y ago
I've supported 3 different models over the years with inference implemented in SQL. First one I inherited, loved it so much that I implemented it twice again. Amazingly fast for TBs of data and no waiting on other tech teams. That tool
53.
▲
by
adeelk93
4y ago
Exactly - and the challenge with that is the data is sometimes only available on BBG. So by not using BBG, you’re missing data, and are at a disadvantage. This is a data problem, not a software problem. The solution isn’t open source, it’s
54.
▲
by
adeelk93
4y ago
why text-davinci and not code-davinci?
55.
▲
by
adeelk93
4y ago
I think there will be a place for AI without limits - but a search engine isn’t that place
56.
▲
by
adeelk93
4y ago
In a war of attrition, yes, but this is a one-off. It’s not purely PR, I am sure the primary motivation is actually counterintelligence, which is why they are recovering it over shallow water.
57.
▲
by
adeelk93
4y ago
It was the size of 195 Goodyear blimps, and bigger than a football field. I imagine bullets would do nothing to it. https://www.forbes.com/sites/davidhambling/2023/02/03/bustin...
58.
▲
by
adeelk93
4y ago
For the pilots, it was - the US achieved air superiority quite quickly. Whatever point you’re trying to make, you lost it along the way. The outcome of of the afghan war was not ideal, but that was not caused by a lack of pilot training, th
59.
▲
by
adeelk93
4y ago
And if you want/need to take it further, I love this package: https://github.com/calogica/dbt-expectations
60.
▲
by
adeelk93
4y ago
Referring to it as a “prank” is what puts it over the top for me
More ›