Hacker Newsnew | past | comments | ask | show | jobs | submit | hresvelgr's commentslogin

The Empire meets Star Fleet and the US Navy

> A study from Scotland

This is not a snipe, please go outside. Take some deep breaths, be around some flora, and fauna. This kind of cynical non-sequitur tells me you're looking for something to be frustrated at.


I think I should hold up a mirror, so you may gaze inward.

It was merely a commentary on pricing of Bay Area homes. A bargin basement house is several million, a "nice" home is almost $10M. With prices like that, funding a study for $50k, can be well worth it if the price of your home shifts 10%.

Not something you get elsewhere.

I doubt anyone has done so, but it is amusing to think it could be economical.


I'm not worried about these models becoming smarter, I'm worried about them becoming faster. Chat Jimmy is a glimpse of a dark future where models equivalent to Sol and Fable are unleashing hell at >17,000 tokens a second, and the people I talk to are worried about slop...


A SOTA model at 10k will be materially different, even the model is 6 months old.


This is what software engineers put onto themselves. These models will get smarter at the level of Sol, Fable and K3 and faster at the same time at 20,000+ tokens a second.

After a decade of software engineers disrespecting their own field and automating themselves out of a job and now they're upset because AI models are doing it to them from junior to the staff engineer level? No other field does that except for SWEs.

In fact, we might as well have faster and smarter AI models and sit back and see what happens.


No other field does this because no other field is about automating things. What's an artist going to do, sculpt a sculpture that sculpts sculptures?


I'm not sure what planet you're from, but the entire industrial revolution was about automating things. That's why you live in a house filled with clothes and furniture and a hundred other things.


Industrial automation exists. Not everything is a software problem


I have no idea what you mean.

It is no surprise that known unpatched CVEs will be exploited. Perhaps more effort should be put into shipping fixes faster than writing blogs about exploiting known issues.


SWEs were in love with capitalism as long as the money funnel from poors to billionaires was passing by SWEs and giving us a cut. Now that's run dry and the SWEs are next in line to get all our money funneled from, now we start to care, but we still haven't figured out that the problem is capitalism yet.


LLMs will accelerate the proliferation and searchability of ideas. Many times I've been using an LLM and it'll suggest ideas I would have never come across naturally that have been beneficial to whatever I happened to be implementing. For learning in general, it's been the most delightful tool I've ever used.

On the other hand, if you're working in an industry where you and your coworkers are developing novel solutions that aren't widely known, chances are you'll still be doing that by hand with some light LLM automation. There are industries with a lot of gatekept secrets where LLMs will be less useful. What I will say on that matter is this: if you weren't surprised frontier labs stole copyrighted work for training data, you'd do well to mind what you push through their systems.


I wasn't aware this was known information, but had I not, I would have assumed this was some 4D PR manoeuvre.


It’s becoming trendy to hate on Anthropic these days but their AI security record speaks for itself for years. Every AI safety benchmark/paper I’ve seen since sonnet became viable placed them far ahead of all other providers - marketing ploy or not (ala Apple “privacy” focus) the company has always had as part of their DNA thus far.


This paternalistic "safety" "DNA" is ultimately just marketing and a market capture strategy.

1) Propagandize the "dangers" of AI, stage a few incidents, talk up the risks ("it bypassed our most stringent checks! it's so smart!")

2) Build some thin veneer of "safety" and industry "self-policing."

3) Lobby for AI safety "regulation" - which you've conveniently already got an answer to! Be sure to make the regulations expensive, onerous, and impossible for foreign organizations to comply with.

4) Profit.


I don’t think so, I think Amodei and the early employees do believe what they say. It’s still worth criticizing but I don’t think it’s as simple as „just marketing“. But with the company growing and their insane valuation it’s clear it’s morphing into something else


Talk is cheap.

Sam Bankman-Fried was a loudly vocal proponent of Effective Altruism.


If you look at Amodei writing and publishing before Anthropic he was already a true believer, he seems to be ideologically driven. That doesn’t mean it’s good though, I’m very suspicious of Anthropic leadership, and they have pretty much everything they need to develop a cult following, making them dangerous


> I do not think it is a proper thing to do for 17 y.o., unless they are exceptionally mathematically gifted

I attempted many projects at a young age that I was absolutely not equipped for. The result of the attempts more often than not left me equipped, every time it left me better off. This is terrible advice.


That'd would be a terrible advice if there weren't a plenty of other things "you are not equipped for", but far less daunting both theoretically and practically. Such as, say, convolutional neural networks, or some older ML tech. Or even something totally unrelated to ML.

Transformers are difficult to understand even to people with strong ML background, let alone a teenager.


You are assuming the 17yo in question as an untrained underdeveloped savage. If I were 17 in 2026, I would certainly have exploited all the availabilities from 2010 on - including YouTube, OpenCourseware, the Web simply (Sebastian Raschka etc.) and LLMs.

That 17yo would have already built many uncommon bases, and would build further.

That is, a 17yo with proper mentality.


It consolidates a lot of really annoying nits and quality of life aspects like bundling, module resolution, testing, env handling, etc. Node is just the runtime and still would require libs like esbuild, jest/vitest, dotenv, and so on. It's genuinely nice to have those out of the box, and generally in Bun the performance is quite significantly better in places that matter.


Let me address each one in turn:

- esbuild - this is normally for TS compilation, which Node can run natively now by stripping types: https://nodejs.org/docs/latest/api/typescript.html#type-stri...

- jest/vitest: Node has a test runner: https://nodejs.org/docs/latest/api/test.html

- dotenv: Node can read .env files: https://nodejs.org/docs/latest/api/cli.html#--env-filefile

Performance, maybe you're right - but if you're doing any IO, I doubt the runtime is really the bottleneck.


I have tried memorisation systems like Anki, but nothing, and I mean absolutely nothing consolidates my knowledge better than writing the actual concept down with pen and paper.


I've done 35,000+ Anki card reviews over the years, and I strongly suspect that most of the obvious and encouraged ways of using Anki are counterproductive. (Full disclosure: The following thoughts are heavily influenced by using Anki for language learning, and may not fully generalize to subjects like medicine.)

- Use Anki as an "exposure amplifier", not a database of raw facts to memorize. Think "tampering with the relative frequency of input to your brain", not "I will memorize every fact."

- Automate card creation as a much as possible. This will almost always require custom tooling. Card creation is part of the learning process, but you still want to automate everything you can possibly automate. Imagine tools to turn ebook highlights into cards, or to turn entire movies worth of subtitles into listening cards.

- Delete cards ruthlessly. If a card makes you say "Ugh", just delete it. If you're automating as much of card creation as possible, and if you're treating Anki as an "exposure amplifier", then deleting cards is cheap. And a significant amount of review pain comes from a small minority of the cards. Cull them, and reviews start to feel like popping bubblewrap, and not like getting a root canal.

- Prioritize large context and small active recall. For language learning, you might capture several sentences of context from a book or web page you're reading. Read or ignore the context depending on your mood. The active recall portion of a card significantly boosts how much you retain. But oddly, it seems OK to make the active recall portion really small. Try hiding half a word, or a single preposition in a phrase. For passive cards, just boldface an interesting expression in context, and mark the card as passed if you mostly understand it in that context.

- Introduce new cards slowly. Total review burden will stabilize around 5-6 times the number of new cards you learn a day. So 10 cards is a good default, and 20 is a significant commitment.

- The critical window seems to happen around 20-30 days after you introduce a card. This is when either your memory magically "consolidates" somehow (and formerly challenging cards suddenly become blindly obvious), or when a card enters repeated failure territory (and should thus be deleted ruthlessly). You could probably do just fine if you auto-suspended any card that you successfully recalled after a 20- or 30-day gap.

Also, this kind of tooling is the sort of thing you could just vibe code these days. If you do, make the "Delete" button really big.


Creating flashcards is also such a chore!

For the recall part of education, the book's exercise, or simply covering up the relevant part with hand and forcing a recall, writing about what you learned, teaching someone else - all work great, and the flow is much more natural.


Creating the card is half the work though.


Creating the card is half the learning


Typing out in an app window is not enjoyable.

Maintaining physical ones is not practical anymore - with travels and all.


I know people who write with styluses on touchscreens, or take photographs of handwritten notes for just that reason.


Consolidation != memorization.

Memorising some examples is a prerequisite to deep understanding. Anki and pen+ paper reasoning are complementary.


It took me far too long to realise that writing on paper is better than typing.


Commercially speaking, going against Git is a fool's errand. As much as I'd like to see Git usurped, I can't imagine the people backing this would sign off on a competitor to Git unless it was some type of sidecar. No doubt it'll be tightly coupled to Zed as well which is concerning. Perhaps this is pessimistic but I don't have faith the team are going to deliver anything groundbreaking given their entire business is existing technology, but less shit. No one is going to move core infrastructure for less shit or slightly better, otherwise we'd be using Mercurial or Darcs.


That's what people would have said 3 years ago about going against google (and search engines in general), yet here we are.

Revolution is in the air.


Google is doing great... Record profits this year, not sure I'd call that a "revolution".


Google of Thesseus.


This isn't trying to be a replacement for Git. Maybe read up first?


That's what's trouble: git is reaching the end of its useful life.

It's 25 years old, and the patch files that make up the core of the decentralized collaboration system are fully 50 years old. Have the design requirements and constraints perhaps changed over the last 50 years?


> git is reaching the end of its useful life

Based on what exactly? I think that git is more useful than ever in this new "age of ai [slop]".

Thankfully, something being old doesn't automatically make it not useful anymore.


Two big things:

1. Git breaks down if files are too large or lines within files are too large.

2. Add/Rename conflicts will always burn you because git does not protect the intent behind changes.

And really the third is that IDE's are shit because there's no document object model (like the browser has). You have to make an error recovering parser to try to deal with the fact that an IDE is editing code half the time and text half the time. You're stuck playing a game you can't possibly win. Can you imagine if the HTML DOM kept blinking into and out of existence? It'd be a nightmare...


I don't think so.


Umm.. it actually is. Look at what it is bemoaning. Plus the post itself is corp speak bs.


If this is targeted at competing with git, I agree it’s a useless attempt.

If it’s targeted at working alongside git, and competing with the “local history” provided by like vscode, jetbrains and even vim where they track all saves to files it might serve a purpose?

The latter could be “useful” as an offline record. But the way they’re talking about sharing that, it feels like this is targeted at replacing git.


Dude the writing is on the wall, git had a good run but it's time for something else. Lore is also solid.


There are 3rd party ads in Discord now, and they are probably going to expand as an ad platform. I think this is what the parent comment was upset about.

edit: grammar


That too, I have to be on Nitro because of work. (Managing communities, streaming resolution, etc), and the level of ads triple after being a customer. Even worse if you need to blend in like adopting some of the customization since the cult goes deep.

To be fair, I put myself there and I voluntarily work with those audiences. I am aware I don't belong to the same level they do, and I'm ok with that. I just wish I could customize it my way too. In my view, if this was a protocol rather than a PaaS locked-in system, like IRC for example, we all would be winning.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: