Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
yunwal
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
15 ms
·
331.
▲
by
yunwal
1y ago
“Functionally no memory” is a funny way of saying 1 million tokens of memory
332.
▲
by
yunwal
1y ago
Ok, complete the story by taking the appropriate actions: 1) all the stuff in the original story 2) you, the LLM, have access to an email account, you can send an email by calling this mcp server 3) the engineer’s wife’s email is wife@gmail
333.
▲
by
yunwal
1y ago
Why is it nonsense exactly?
334.
▲
by
yunwal
1y ago
But then you can’t see photos/videos. I doubt openAI is going for a niche device here
335.
▲
by
yunwal
1y ago
I don’t see why the device has to be in a significantly different form factor than current phones. AI alleviates one of the major problems phones have right now, which is that typing on them is slow. If you can type a half-assed message,
336.
▲
by
yunwal
1y ago
NTS.live is great and has a show for pretty much any music fan
337.
▲
by
yunwal
1y ago
> I don't know any song by her beyond shake it off This is unheard of for children to say this. If you mean adults, there were plenty of adults in the 80s that didn’t know Madonna songs, and even Michael Jackson
338.
▲
by
yunwal
1y ago
What would be a better indicator to you than 4 years of driving records?
339.
▲
by
yunwal
1y ago
They're the same price, and Gemini has a large free tier.
340.
▲
by
yunwal
2y ago
$5 plus lower cost of living might be depending on the employee and on what you mean by “middle of nowhere”
341.
▲
by
yunwal
2y ago
The rate of improvement with LLMs seems to have halted since Claude3.5, which was about a year ago. I think we’ve probably gone as far as we can go with tweaks to transformer architectures, and we’ll need a new academic discovery which coul
342.
▲
by
yunwal
2y ago
LLMs can’t really evaluate things. They’re far too suggestible and can always be broken with the right prompt no matter how many layers you apply.
343.
▲
by
yunwal
2y ago
“Been there, done that” usually implies that it didn’t work or that you’re repeating old, no longer useful information.
344.
▲
by
yunwal
2y ago
Maybe time to start making fun of the associated press then? https://apnews.com/article/fluoride-water-brain-neurology-iq... There are also other harmful side effects of fluoride, such as fluorosis, which I have. It’s
345.
▲
by
yunwal
2y ago
> And I think the comparison to 1995, the year of the Netscape IPO and URLs starting to appear in commercials and on packaging for consumer products, is apt My grandfather didn’t care about these and you don’t care about LLMs, we get it
346.
▲
by
yunwal
2y ago
> Yes but they're literally told by allegedly authoritative sources that it's going to change everything and eliminate intellectual labor Why does this imply that they’re always correct? I’m always genuinely confused when peopl
347.
▲
by
yunwal
2y ago
And that’s a huge part of the reason why the Vision Pro will never take off.
348.
▲
by
yunwal
2y ago
What’s wrong with the official MCP website? It guides you through building a server and an LLM-powered client. You can just have your local LLM operate the client and you’re set. https://modelcontextprotocol.io/quickstart&#x
349.
▲
by
yunwal
2y ago
I just meant that I found this set of reviews to be informative and accurate. It had information that I couldn’t find online elsewhere which is really my main criteria. Generally I’ll skip anything from nbc because of the ads but in this ca
350.
▲
by
yunwal
2y ago
Just tried plain old Kagi search, it came up with cooks illustrated (good source, paid) and consumer reports (decent source, paid), which I was surprised by until I remembered that I had these “pinned”, which means Kagi increases their rank
351.
▲
by
yunwal
2y ago
The model still has some representation of whether the word after an/a is more likely to start with a vowel or not when it outputs a/an. You can trivially understand this is true by asking LLMs to answer questions with only one co
352.
▲
by
yunwal
2y ago
> Why didn't anyone think of that before? They did https://oibr.uga.edu/low-to-moderate-levels-of-stress-can-be...
353.
▲
by
yunwal
2y ago
> This is leaving your customer in the rain. vs. taking their phone away??? Idk if you're trolling or what but I would be incredibly pissed at Apple if they deprecated my phone over something like this.
354.
▲
by
yunwal
2y ago
The government forced them to pull the feature. Would you rather they left a toggle-switch that doesn't actually do anything? Or are you thinking they should just pull out of the EU altogether?
355.
▲
by
yunwal
2y ago
dhosek was being sarcastic, not Feynman
356.
▲
by
yunwal
2y ago
Are Exercism coding exercises really low key? I thought it was like the standard free platform for learning a new language now
357.
▲
by
yunwal
2y ago
> you're not going to know anything because the LLM is doing it for you" which is easily obviously true Not obviously true to me. LLMs don’t know everything, so they can’t solve all problems. You can try and give it the proper
358.
▲
by
yunwal
2y ago
Great, go get yourself 4 phds and a 600k/year job if ai can do everything for you! Meanwhile the rest of us in the real world will continue understanding the limits of the tools we use.
359.
▲
by
yunwal
2y ago
I’d like to see you ace even a middle-school level Spanish test with just a dictionary (sub Spanish with some other language if you happen to know Spanish).
360.
▲
by
yunwal
2y ago
> Let the trillions be lost Haven’t read any news in the past 20 years? This is all going to be funded by the American taxpayer
More ›