kentm

7 ms·

91.
by kentm 10mo ago
> did I write that paragraph? No. My kid wrote a note to me chock full of spelling and grammar mistakes. That has more emotional impact than if he'd spent the same amount of time running it through an AI. It doesn't matter h
92.
by kentm 10mo ago
Do you think that spam letters are generally considered to be a good use of resources?
93.
by kentm 10mo ago
They do the same thing with the i in “doing” in another post. It seems like just a typo this person sometimes makes.
94.
by kentm 10mo ago
> Did the models from 2 years ago produce more bugs, fewer bugs or the same bugs as today's models? Is anyone actually tracking that with a methodology not prone to fine-tuning? Specifically, I know a lot of the tests have the prob
95.
by kentm 10mo ago
> These tools are here to stay. I don't think that these specific tools are here to stay. Categorically, yes, but I expect there to be big changes in interfaces and how they work in the next five years. I don't think spin up
96.
by kentm 10mo ago
B-trees too but also bucketing for formats like delta lake or iceberg, where having ids that cluster will reduce the number of files you need to update.
97.
by kentm 10mo ago
> You can argue that, but then what is its purpose? The purpose is to reduce randomness while still preserving probability of uniqueness. UUIDv4 come with performance issues when used to bucket data for updates, such as when there used a
98.
by kentm 10mo ago
> How come ad supported TV existed for decades without destroying children's mental health? I would argue that it did, we just did a poor job of measuring it. Anecdotally, during my childhood I moved from a place that had very littl
99.
by kentm 10mo ago
Respect the robots.txt and don’t do it?
100.
by kentm 10mo ago
He mentions living through the dialog fast and says that the takeaway is that people don’t care about the story. However, I wonder if his players just read very fast. I grew up on JRPGs and read through dialog quickly, to the point that peo
101.
by kentm 10mo ago
A big task my team did had measured accuracy in the mid 80% FWIW. I think the line of thought in this thread is broadly correct. The most value I’ve seen in AI is problems where the cost of being wrong is low and it’s easy to verify the out
102.
by kentm 10mo ago
> have you considered that you haven't used the tools correctly or effectively? The problem is that this comes off just as tone-deaf as "you're holding it wrong." In my experience, when people promote AI, its sold a
103.
by kentm 10mo ago
> where some LLM will try to parse that, fail miserably, and then shunt you to an actual person. If you're lucky. I've had LLMs that just repeatedly hang up on me when they obviously hit a dead end.
104.
by kentm 10mo ago
Personally, EMR has never shaken off the "scrappy" feeling (sometimes it feels OK if you're using Spark), and it feels even more neglected recently as they seem to want you on AWS Glue or Athena. LakeFormation is... a thing
105.
by kentm 10mo ago
This person wrote a blog post admitting to tone-deafness in cheerleading AI and talking about all the ways AI hype has negatively impacted peoples' work environment. But then they wrap up by concluding that its the anti-AI people that
106.
by kentm 10mo ago
> It's not helping that in the last 10 years a culture of job-hopping has taken over the tech industry. Average tenure at tech companies is often ~2 years and after that people job hop to increase compensation. I've started vie
107.
by kentm 10mo ago
I'd add SQS to the solid category. But yes, the less of a core building block the specific service is (or widely used internally in Amazon), the more likely you are to run into significant issues.
108.
by kentm 11mo ago
Art, writing, and communication is about humans connecting with each other and trying to come to mutual understanding. Exploring the human condition. If I’m engaging with an AI instead of a person, is there a point? There’s an argument that
109.
by kentm 11mo ago
Yeah, Voyager 1 was launched on a Titan IIIE. I don't really want to do the delta v calculations, but if we look at mass to LEO as a rough proxy, Titan IIIE does 15,400 kg and the Falcon Heavy does around 50,000 kg (with re-use). New
110.
by kentm 11mo ago
The charitable answer would be that they admire self reflection and try to engage i it themselves.
111.
by kentm 11mo ago
Likely not to their face, but I’ve absolutely had tons of people use it when talking to me as a slur for other people. Because they assumed I was like minded. It’s absolutely used as a slur in conversation, yes. I mean you have the Republic
112.
by kentm 11mo ago
In my opinion, the problem is that journalists in general used “both sides” rhetoric where it wasn’t warranted to avoid accusations of bias. It feels that nuance is used out of cowardice more often than not. There’s also the fact that not a
113.
by kentm 11mo ago
Yep agree with this. Nothing is more infuriating than someone Kramering into a space trying “to help” without spending any time or effort trying to understand that space. They should have understood the guidelines before turning on their ma
114.
by kentm 11mo ago
Turning off the machine translation and reverting all the changes it made seems pretty actionable to me. They can turn it back on when issues are addressed.
115.
by kentm 11mo ago
Don’t use passive voice in an apology. “We’re sorry that we made the change without consulting your team or considering your circumstances.” The change did not fall out of thin air. It was something they did. If they do not own it explicitl
116.
by kentm 11mo ago
Correct. Companies are absolutely falling over themselves to replace high quality human translations with lower quality machine translation. I’m not sure how a hacker news poster could miss this trend.
117.
by kentm 11mo ago
Different parts of the analytical stack have different performance requirements and characteristics. Maybe none of your stack needs it and so you never need Snowflake at all. More likely, you don't need Snowflake to process queries fr
118.
by kentm 11mo ago
It's not going to scale as well as Snowflake, but it gets you into an Iceberg ecosystem which Snowflake can ingest and process at scale. Analytical data systems are typically trending to heterogenous compute with a shared storage back
119.
by kentm 11mo ago
A talk I went to made the point that LLMs don't sometimes hallucinate. They always hallucinate -- its what they're made to do. Usually those hallucinations align with reality in some way, but sometimes they don't. I always
120.
by kentm 1y ago
Frankly I’d be incredibly exited if the next Apple OS update was “No new major featurs. Bug fixes, perf optimization, and minor ergonomic improvements only”.

More ›