HNHacker News
TopNewBestAskShowJobs

thequadehunter

187 karma · joined February 24, 2023

submissionscomments
thequadehunter··on LangChain Announces 10M Seed Round
Fixie got a 17M seed so I guess not LOL
thequadehunter··on LangChain Announces 10M Seed Round
What do you suspect could go wrong? Not attacking, just curious.
thequadehunter··on LangChain Announces 10M Seed Round
Another thing people seem to forget is that Langchain can use LLMs that aren't made by OpenAI.

If OpenAI goes under, or a great open-source model comes onto the scene, Langchain can still do its thing.

thequadehunter··on Show HN: We are building an open-source IDE powered by AI
As an infra dude it saves me loads of time. Most of the code I write is stuff where I already know what needs to happen, I already know how the process looks un-automated, and I'm not going to learn a whole lot writing more boilerplate, or learning the names of all the endpoints for a dumb api and parsing the text.

It's still mental work. You still have to read the code over and edit stuff.

It helps me save lots of time so I can do quality things like comment on hn during work..

thequadehunter··on LangChain Announces 10M Seed Round
I was about to rant about the documentation, but I just checked and it seems to have improved a lot.
thequadehunter··on LangChain Announces 10M Seed Round
It's unfortunate that prettymuch all shiny AI things have horrible documentation. I see a lot of misinformation in non-researcher AI circles and I feel like it stems from that sometimes.
thequadehunter··on Grand jury votes to indict Donald Trump
Weirdly enough, I've been far more interested in local politics lately because I can see stuff actually get done that improves my life.
thequadehunter··on Grand jury votes to indict Donald Trump
I don't really understand this. What evidence is there that this is the case?
thequadehunter··on Grand jury votes to indict Donald Trump
The fact that you're factoring in how popular he polls to determine whether this prosecution is worth it cements in my mind that it is 100% worth it.

Popularity should be the last factor in whether someone is prosecuted for a crime.

thequadehunter··on Cerebras-GPT vs. LLaMA AI Model Performance Comparison
Apparently it can.

https://writings.stephenwolfram.com/2023/03/chatgpt-gets-its...

thequadehunter··on Cerebras-GPT vs. LLaMA AI Model Performance Comparison
Oh, I see. It's about a cap in baseline ability rather than whether it's technically possible to give it the extension?
thequadehunter··on Cerebras-GPT vs. LLaMA AI Model Performance Comparison
I've been wondering...does it even matter at this point when you can just pull out a calculator in langchain?

Also side note...I went to high school with the author and hadn't heard about him for a while lol. That was a neat surprise.

thequadehunter··on Retrieval in LangChain
That's funny...I actually just bought your book because I have a bone to pick with the documentation. It's good at explaining stuff, but I feel it fails to show how some of the parts work together. It would be immediately more obvious if there were just full examples rather than partial ones.

Thanks for writing this.

thequadehunter··on Show HN: ChatGPT Plugins are a security nightmare
Ah, thanks. Makes sense.
thequadehunter··on Show HN: ChatGPT Plugins are a security nightmare
So...in the pirate example the comment said to talk like a pirate, right? Is the example comment where it searches for a keyword a different example?

I'm just really confused why the image says to search for a keyword, and then the LLM comes back talking like a pirate.

thequadehunter··on Post-GPT Computing
All your points I agree with. I didn't say it was a good system. I've lived in Japan and I loved the government health insurance. One of my biggest annoyances about being back in the US is how I get my insurance through my job, so now every time I hop jobs I have to do the stupid math and figure out the new employer's health crap.

I'm just saying that I see these weird takes about Americans paying like 10k for a single visit when that's just not the reality for most people. Only 8.6% of Americans were uninsured in 2020. Most insurance plans have a cap that doesn't even allow you to pay more than like 2-10k in an entire year.

Is it more expensive? Duh. It's just that there's more nuance to the issue. If I wasn't American (or I was an American that didn't understand how insurance actually works) I might have read that and believed that the poor Americans are all living in hell.

So yeah...we can talk about the issue because it's a big deal. In fact, I think it's the #1 biggest problem with American society right now and it's maddening that I never see anyone focusing on it. However, I don't like to see this kind of hyperbole from OP because people turn their brains off and stop looking at the facts.

thequadehunter··on March 20 ChatGPT outage: Here’s what happened
Yeah but again, this is the fastest growing app in history and it uses way more compute than your standard webapp, and basically delivers all functionality from a single service that handles that load. I can see why there would be some growing pains.
thequadehunter··on Hello Dolly: Democratizing the magic of ChatGPT with open models
Lol. This is classic ML crap. Files with no documentation, no links, multiple files with the same-ish name but no explanation for which one is what.
thequadehunter··on Post-GPT Computing
I can't see the first case ever happening. You'd need a whooooole lot of trust in AI systems to have it write all the code.

As for the latter...I'd say GPT has increased my productivity and therefore allowed me to focus on the more interesting aspects of my work, rather than writing annoying boilerplate code and doing boring tasks where I don't learn anything. I almost never write my own boilerplate anymore.

More productivity doesn't necesarily mean more work. It does mean more focus on interesting work.

thequadehunter··on Post-GPT Computing
TBH, the most useful thing it does for me is write my boilerplate code and tedious statements. That's infinitely more useful than the knowledge stuff.

The other day I was working with the Cisco Meraki API...I knew exactly what the script needed to do, but the calls were tedious and I didn't feel like learning the names of all the JSON columns, so I just had ChatGPT do it. I had to fix a couple mistakes, but the 20 minutes it took was better than having to read all the documentation.

thequadehunter··on Post-GPT Computing
The other thing is that we'd need insane levels of trust in the AI to have it doing all these jobs 100%.

Like, I could technically have a newbie running commands on a production router for a script that I wrote out...but even if I let them do that there's no way I wouldn't at least supervise. I don't think most companies are even remotely comfortable with the idea of having an AI system running code on their systems no matter how smart it is.

thequadehunter··on Post-GPT Computing
Just FYI for non-Americans...most Americans are insured and doctors visits don't cost 10k. Major surgeries might, but your insurance usually caps at a certain number in the 2-8k range for the whole year.

Not saying the system isn't bad, but 10k for a doctor's visit is kind of a stretch...

thequadehunter··on OpenAI tech gives Microsoft's Bing a boost in search battle with Google
This has been my thing too. I work in Networking and sometimes I see interesting stuff, but between cisco's beefy docs and obscure blogs I have no idea what the actual use case for it, or how the SD-WAN magic wand does the thing. ChatGPT will get me to verifiable information so fast, and give great in-context examples. It's renewed my passion for the field just when it was getting dull.
thequadehunter··on March 20 ChatGPT outage: Here’s what happened
To be fair, they accidently made a game-changing breakthrough that gained millions of users overnight, and I don't think they were ready for it.

Before chatgpt, most normal people had never heard of OpenAI. Their flagship product was basically an API that only programmers could make useful.

Team leaders at OpenAI have stated that they were not expecting the success, let alone the highest adoption rate for any product in history. In their minds, it was just a cleaned-up version of a 2-year old product. It was billed as a research preview.

So, all of a sudden you go from hiring mostly researchers because you only have to maintain an API and some mid-traffic web infra, to suddenly having the fastest growing web product in history and having to scale up as fast as you can. Keep in mind that they didn't get backing from Microsoft until January 23, 2023-- that was only 2 months ago.

I'd say we should cut them some slack.

thequadehunter··on March 20 ChatGPT outage: Here’s what happened
Personally, I think it was partially a virtue signal to show that they use open source software and collaborate with the maintainers.
thequadehunter··on Who is still inside the metaverse?
VRchat. It's the best or worst VR experience depending on who you are.
thequadehunter··on Who is still inside the metaverse?
I never understood the metaverse thing. Was Zuck unaware of vrchat? I was always confused why they thought users of a niche market geared towards hackers and nerds would gravitate towards the soul-less corporate solution as opposed to the existing one.
thequadehunter··on ChatGPT is down
It's actually been down for like over 12 hours. Tried to use it last night.
thequadehunter··on On the left is GPT-3.5. On the right is GPT-4
What would it take for you to consider it alive?

Or, to simplify...why is a dog "alive" but a chatbot with a cognition loop is not?

thequadehunter··on On the left is GPT-3.5. On the right is GPT-4
This is what I'm talking about. Everybody in the comments is doing that thing where "it's not sentient bro, it's just an app that responds to input!" but that's so beside the point. Humans parrot points they heard on HN all the time too. The question is more about how can you be sure it's not sentient based on how it acts, and how can you be sure that someone can't make it sentient by giving it something like a cognition loop?

If you had an LLM that had unlimited chat storage, and gave it multiple containers for things like memories, idle thoughts, daydreams, aspirations, etc, and you had a loop for each of those where it reads information and reasons through it to form those thoughts, then stores those thoughts in certain containers and brings it up in conversation...would you feel bad if you called it a name and it reacted with emotional text? It's still "just an llm" with a bunch of overhead.

← PreviousPage 2 of 3Next →