HNHacker News
TopNewBestAskShowJobs

pera

11,937 karma · joined June 5, 2012

submissionscomments
pera··on Andrej Karpathy: Software in the era of AI [video]
Is it possible to vibe code NFT smart contracts with Software 3.0?
pera··on Andrej Karpathy's talk on the future of the industry
Is "Software 3.0" somehow related to "Web 3.0"?
pera··on Generative AI coding tools and agents do not work for me
That would be a reasonable thing to do, unfortunately this doesn't always happen. Say for example that your company is quite behind schedule and decides to pay some cheap contractors to work on anything that doesn't require domain expertise: In 2025 these cheap contractors will 100% vibe code their way through their assigned tickets. They will open PRs that look "nearly there" and basically hope for all green checks in your CI/CD pipeline. If that doesn't happen then they will try to bruteforce^W vibe code the PR for a couple of hours. If it still doesn't pass then claim that the PR is ready but there is something wrong for example with an external component which they can't touch due to contractual reasons...

One of the most bizarre experiences I have had over this past year was dealing with a developer who would screen share a ChatGPT session where they were trying to generate a test payload with a given schema, getting something that didn't pass schema validation, and then immediately telling me that there must be a bug in the validator (from Apache foundation). I was truly out of words.

pera··on Meta's Llama 3.1 can recall 42 percent of the first Harry Potter book
It's not literally the title of the article, nor the premise of its first paragraph, but since this was your interpretation I wonder if there is a misunderstanding around the term "piracy", which I believe is normally defined as the unauthorized reproduction of works, not a synonym for copyright infringement, which is a more broad concept.
pera··on Meta's Llama 3.1 can recall 42 percent of the first Harry Potter book
> let's not pretend that an LLM that autocompletes a couple lines from harry potter with 50% accuracy is some massive new avenue to piracy

No one is claiming this.

The corporations developing LLMs are doing so by sampling media without their owners' permission and arguing this is protected by US fair use laws, which is incorrect - as the late AI researcher Suchir Balaji explained in this other article:

https://suchir.net/fair_use.html

pera··on I'm the CTO of Palantir. Today I Join the Army
You are not a real corporatocracy until you don't have military ranks just for your C-suite pals
pera··on Ask HN: What Happened to the Apple Vision Pro?
Ah yeah I see they are still being sold online, I thought they were discontinued
pera··on My AI skeptic friends are all nuts
Yeah I don't agree that having more start-ups checking a box saying that at least one of their component uses some form of AI indicates that we are experiencing a surge in new start-ups.

The amount of start-ups getting into YC hasn't really changed YoY: The W23 batch had 282 companies and W24 260.

pera··on My AI skeptic friends are all nuts
Sorry I don't follow, would you mind clarifying your point?
pera··on My AI skeptic friends are all nuts
It's fascinating how over the past year we have had almost daily posts like this one, yet from the outside everything looks exactly the same, isn't that very weird?

Why haven't we seen an explosion of new start-ups, products or features? Why do we still see hundreds of bug tickets on every issue tracking page? Have you noticed anything different on any changelog?

I invite tptacek, or any other chatbot enthusiast around, to publish project metrics and show some actual numbers.

pera··on My AI skeptic friends are all nuts
A flatline also looks locally the same at all points in time.
pera··on My AI skeptic friends are all nuts
It's a bit more than a metaphor :) during the California gold rush there was this guy named Sam Brannan who sold shovels and other tools to miners, and made a fortune from it (he is often referred to as California's first millionaire). He also had a newspaper at the time, the California Star, which as you can imagine was used to promote the gold rush:

> The excitement and enthusiasm of Gold Washing still continues—increases. (1848)

https://sfmuseum.org/hist6/star.html

https://en.wikipedia.org/wiki/Samuel_Brannan

pera··on Cloudlflare builds OAuth with Claude and publishes all the prompts
That's interesting. My experience with Sonnet 3.7 early this year was pretty poor: It simply couldn't reach the correct solution alone, even when explaining the issues explicitly. The proposed invalid solution was not too far from the correct one, so you could fix it manually if you knew what you were doing, but then the way the code was structured was not something that I would like to maintain in a real project. All this on top of the usual UX issues like hallucinated APIs. The experience refactoring was even worse.

I guess your mileage is highly dependent on the domain of your problem? In my case was GIS by the way

pera··on The ‘white-collar bloodbath’ is all part of the AI hype machine
For those outside of the US:

https://en.m.wikipedia.org/wiki/Second_inauguration_of_Donal...

pera··on The ‘white-collar bloodbath’ is all part of the AI hype machine
Bullshit: Chatbots are not failing to demonstrate a tangible increase in companies' output because of regulations and IP law, they are failing because they are still not good for the job.

LLMs only exist because the companies developing them are so ridiculously powerful that can completely ignore the rule of law, or if necessary even change it (as they are currently trying to do here in Europe).

Remember we are talking about a technology created by torrenting 82 TB of pirated books, and that's just one single example.

"Steal all the users, steal all the music" and then lawyer up, as Eric Schmidt said at Stanford a few months ago.

pera··on What's working for YC companies since the AI boom
You are being downvoted but I think you are correct: I'm not sure I could name a single company that came out from YC in the past 10 years, which makes think that YC was essentially pg, and I'm not a fan of him but he was obviously a very talented business person and had a really good eye for startups.

YC's leadership nowadays seem to lack of any kind of vision and just follows whatever shiny tech is currently in vogue, which is a huge red flag for accelerators.

pera··on Ask HN: Do we need AGI to filter spam?
Would this AGI be so good that even you wouldn't be able to tell whether it's spam or not? That's hard to imagine
pera··on Google is burying the web alive
The corpse was already buried ages ago and his name was Jeeves: What they are doing now is an exhumation and calling it the future.
pera··on Inhibiting 15-PGDH blocks BBB deterioration and protects mice from Alzheimer's
Original title: Inhibiting 15-PGDH blocks blood–brain barrier deterioration and protects mice from Alzheimer’s disease and traumatic brain injury
pera··on Watching AI drive Microsoft employees insane
This happened during LlamaCon while taking about Copilot/LLMs: if the percentages Satya was referring to were for any "auto-generated" code then he was being intentionally misleading.

Besides, you could also say that 100% of code is generated "by software" no?

pera··on Watching AI drive Microsoft employees insane
This is all fun and games until it's your CEO who decides to go "AI first" and starts enforcing "vibe coding" by monitoring LLM API usage...
pera··on MIT paper on AI for materials research found to be fraudulent
This is actually an in-depth analysis on the fraudulent paper and it's much more interesting than the WSJ article. One for dang to decide I guess.
pera··on The Brief Origins of May Day
We have been conditioned to fear the idea of demanding a more fair distribution of wealth: My great grandfather for example was shot dead by a cop during a union strike. Nothing happened after that event, just silence, fear and a deeply traumatized family.

It is very important to remember that many of the things we enjoy today as workers only exist because of the enormous courage of workers from the 19th and early 20th century who fought really hard for a better future.

pera··on "AI-first" is the new Return To Office
It may have been stated somewhere in their T&C, but just to be clear: I did not explicitly consent for this.
pera··on "AI-first" is the new Return To Office
Not from my point of view: Last year I paid for a product from a shop that was using Shopify, you would think that given that this was a one time transaction, once the payment has been processed they wouldn't keep storing your PII forever but guess what? A couple of months ago I was going to buy a product from a totally unrelated shop also using Shopify and as soon as I typed my email address they sent me an OTP to my phone to autocomplete all my personal details. So paying for a product equals to creating an account with Shopify.

This is incredibly shady and I wonder if it's even legal here in Europe.

pera··on Reports of widespread power cuts in Spain and Portugal
Parts of France are also affected according to Sky:

https://news.sky.com/story/large-parts-of-spain-and-portugal...

pera··on Ask HN: My CEO wants to go hard on AI. What do I do?
I mean yeah but in a way it's kinda like a legalised pyramid scheme
pera··on Ask HN: My CEO wants to go hard on AI. What do I do?
> The argument seems to be that they've realized the only way to achieve the next round of funding is to be "AI-first".

This is the argument for virtually every company going AI-first: it's just a way to scam VCs.

It seems crazy but it really works.

pera··on Whistleblower: DOGE Siphoned NLRB Case Data
Is Graham still involved in HN?

While it's true that he has spoken against Trump many times, Garry Tan is very close to Thiel, Musk, and the MAGA movement in general. Didn't he recently show support for DOGE as well?

Of course, YC is more than its current CEO and hopefully this doesn't affect the moderation of this website :)

pera··on Eric Schmidt Tells Congress That 99% of All Electricity Will Be Used to Power AI
Original title was too long: Former Google CEO Tells Congress That 99 Percent of All Electricity Will Be Used to Power Superintelligent AI

This title is also inaccurate as what he actually said was "from 3 percent to 99 percent of total generation".

← PreviousPage 6 of 34Next →