HNHacker News
TopNewBestAskShowJobs

sgt101

7,940 karma · joined January 24, 2012

nothing to see here, move along, move along.
submissionscomments
sgt101··on What do wealthy people buy, that ordinary people know nothing about? (2015)
In the spirit of engaging in open conversation to try and broaden my mind and understanding, may I say that I feel your statement "The tax systems doesn’t create wealthy people, it just doesn’t cut their legs out from under them" begs a question: What is it that creates wealthy people?

I think that there are a few things that are needed for lots people to become wealthy (of course, the ability to exact physical violence on others can enable a few to become very very wealthy as well).

1) Property rights that are equally enforced. That you have important friends shouldn't give you the ability to take my business.

2) Liquidity. There needs to be mechanisms that enable investment.

3) Civil infrastructure & some sort of safety net. People who are frightened of what awaits their children, or them in their old age, or of sickness, are reluctant to invest. They will be happy to work themselves half to death, but even if they have enough money, they will be very very risk averse.

Is this what you think, or are there other mechanisms that you have in your mind?

sgt101··on What does the end of mathematics look like?
and instead we have photographers who can document the world at a great volume. My Grandparents had no visual record of their wedding. My wife is a wedding photographer...
sgt101··on What does the end of mathematics look like?
"perhaps ethically questionable"?

I find it hard to think of a more ethically questionable programme!

sgt101··on What do wealthy people buy, that ordinary people know nothing about? (2015)
Well - there is an alternative take. Whisper it... but we don't actually need to have ultra rich people in our society. They serve no purpose that normal rich people don't, and arguably introduce distortions that are profoundly corrosive.
sgt101··on Taking a look at the next generation of telescopes
What are the competitors to the European Telescope?
sgt101··on Launch HN: Nao Labs (YC X25) – Cursor for Data
Any link?
sgt101··on Build real-time knowledge graph for documents with LLM
Maybe ask the LLM to extract facts from the documents as datalog assertions and then use a reasoner/llm tool to answer the question?
sgt101··on Launch HN: Nao Labs (YC X25) – Cursor for Data
Does anyone have any links for more LLM based tools that are aimed at data engineering and data science?
sgt101··on A Survey of AI Agent Protocols
FIPA came after KQML and tried to learn from it and do something better. There were people who didn't think that it had succeeded in that.

So it's debatable I think.

sgt101··on A Survey of AI Agent Protocols
A very old report about FIPA that was rejected from AAMAS.

https://github.com/sgt101/FIPA-Mirror-/blob/main/Experience_...

sgt101··on A Survey of AI Agent Protocols
Doesn't even mention KQML or FIPA - staggeringly ignorant.
sgt101··on Naur's "Programming as Theory Building" and LLMs replacing human programmers
I'm not sure you're right you know. I think that the way that an LLM maintains a conversation is to have the conversational thread fed into an instance of it at every step. You can see this if you do a conversation step by step and then take all of it (including the LLM responses) apart from the final outcome and paste that into a new thread:

https://chatgpt.com/share/6814e827-81cc-8001-a75f-64ed6df5fc...

https://chatgpt.com/share/6814e7fb-f4d0-8001-a503-9c991df832...

if you think about how these things work as services you can see that this makes sense. The model weights are several gb, so caching the model weights for utilisation by a particular customer is impractical. So if the forward pass does update the model then that's instantly discarded, what's retained is the conversational text, and that's the bit that's uploaded to the model on each iteration for a new reply. There are hundreds of requests pinging through the data center where the models are used every second, all of these use the same models.

But if you believe that there is a reasoning process taking place in the text then fair enough.

sgt101··on Naur's "Programming as Theory Building" and LLMs replacing human programmers
>We don't fully understand how we relate words to concepts and meaning ourselves,

This is definitely true.

>but to the extent we do, LLMs are by far the closest implementation of those same ideas in a computer

Well - this is half true but meaningless. I mean - we don't understand so LLM's are as good a bet as anything.

LLMs will confidently tell you that white wine is good with fish, but they have no experience of the taste of wine, or fish, or what it means for one to compliment the other. Humans all know what it's like to have fluid in their mouths, they know the taste of food and the feel of the ground under their feet. LLMs have no experience, they exist crystalised and unchanging in an abstract eternal now, so they literally can't understand anything.

sgt101··on Widespread power outage in Spain and Portugal
> Most of banks worldwide still use COBOL for most (all?) their software infrastructure.

Nahhh - some banks have some parts of the infrastructure in COBOL. Specifically larger retail banks often have their ledgers in COBOL. Most of them want rid and are actively getting rid. Most places have had programs to root COBOL out since before 2000, but there are residual implementations remain. The ledgers are the hardest place to deal with because of the business case as well as the awkwardness. Basically there's not much of an advantage (or at least hasn't been) in modernizing so keeping the thing going has been the option. Now people want to have more flexible core systems so that they can offer more products, although not so sure that customers want this or can consume it. Still - it supports the idea of modernisation so not many people are keen to challenge.

The most common big implementations I come across are in Java.

sgt101··on Widespread power outage in Spain and Portugal
I did a project where we predicted transformer failure (they blow up!) from changes in the oil that they have in them (its insulative properties suddenly change). This was 30 years ago so it's all a bit fuzzy, but the one thing that really stuck with me was the story that the SME that we were working with had about the UK grid teetering on the edge of failure during the "great storm" of 1987. His telling was that they really were unsure if they would get it back at all!
sgt101··on LHC 2025 First Collisions
money
sgt101··on Everything wrong with MCP
These examples seem to be tools built on this tool - which is cool and all, but it's not the equivalent of "a mail order catalogue you can access through your computer", or "a replacement for a travel agent".

What are the applications at the level of Amazon.com, Expedia, or Hacker News?

sgt101··on Everything wrong with MCP
This is a key point - where are the applications?

Normally we have a standard when we have applications, but I am not seeing these yet... perhaps I am blind and mad!

sgt101··on Everything wrong with MCP
This is what torpedoed the first AI Assistant push in the late 1990's and early 2000s (see electric elves as an example). Basically we thought that personal travel assistants and open trading platforms would be cool things, but then discovered that content aggregators a) had a business model and b) could offer vertical integration and bulk buys, & discounts to goods providers and consumers, while also policing the platform. So we got Expedia and Ebay.

There is a more fundamental problem as well. Multi-agent systems require the programmer of the agent to influence the state of the agent in order that the agent can act to influence the states of other agents in the system. This is a very hard programming task.

sgt101··on CERN releases report on the feasibility of a possible Future Circular Collider
>hey plan to provide experiment agnostic and unified framework that is actually unified and user accessible. The field really need something like that, it is usually the pain point of most junior graduate student. The field suck at documentation and keep coherent software and write code in a bad way most of the time. I think they can have much better framework than Fermilab's art [2].

I don't think that I've ever heard of an area of software where folks say "it's got great documentation and the code's really good". In fact many programmers react allergically to any code that they review (including code that they wrote themselves but have forgotten about).

sgt101··on Atari Missile Command Game Built Using AI Gemini 2.5 Pro
I think that the other question is "what do we mean that Gemini 2.5 built it" - was it with a prompt like "build me a missile command game with a reward system that will enable me to buy upgrades"? Or was it interactively with the dev?
sgt101··on Obituary for Cyc
I think logical reasoning - so reasoning about logical problems, especially those with transitive relations like two way implication. A way round this is to get them to write prolog relations and then reason over them... with prolog. This isn't a fail - it's what things like prolog do, and not what things like nns do. If I was asked to solve these problems I would write prolog too.

I think quite a lot of planning.

I think scheduling - I tried something recently and GPT4 wrote python code which worked for very naive cases but then failed at any scale.

Basically though - trusted reasoning. Where you need a precise and correct answer LLM's aren't any good. They fail in the limit. But where you need a generally decent answer they are amazing. You just can't rely on it.

Whereas GOFAI you can, because if you couldn't the community thew it out and said it was impossible!

sgt101··on The "S" in MCP Stands for Security
There are privilege boundaries within which this fundamentally is a problem as well, for example inside banks where this could be used to silently monitor for events that could then be used to trigger frauds or other bad things.

The problem is that it is very hard to see how you can prove this is going to be safely implemented, for example, is it possible to say that your sharepoint or confluence is "safe" in terms of all the content that's in there? I do not think so...

sgt101··on US Administration announces 34% tariffs on China, 20% on EU
It's not like they're the controlling power behind the scenes of the international hedgemon. They're a person commenting on HN. Calm the drama.
sgt101··on Tracing the thoughts of a large language model
>Claude will plan what it will say many words ahead, and write to get to that destination. We show this in the realm of poetry, where it thinks of possible rhyming words in advance and writes the next line to get there. This is powerful evidence that even though models are trained to output one word at a time, they may think on much longer horizons to do so.

Models aren't trained to do next word prediction though - they are trained to do missing word in this text prediction.

sgt101··on Bitter Lesson is about AI agents
There's some challenge there.

- service interactions are pretty complicated - the satisfaction is dependent on a lot of factors - customers mostly give 7/10 mostly because giving nps is so unnatural

sgt101··on Bitter Lesson is about AI agents
I don't get how RL can be applied in a domain where there is no simulator.

So for customer service, to do RL on real customers... well this sounds like it's going to be staggeringly slow and very expensive in terms of peeved customers.

sgt101··on Scallop – A Language for Neurosymbolic Programming
an llm or other nn construct
sgt101··on Scallop – A Language for Neurosymbolic Programming
Let's say you want to program a system that uses a variety of neural nets for reasoning about a problem, and you also want to use more traditional programmatic reasoning - for example to score and rank results. I think this is the kind of language that could allow you to do that.

But - it's time to do the tutorials and try and see.

sgt101··on Scallop – A Language for Neurosymbolic Programming
Thank you.

you seem to be more in the know than me :) Please could you just sketch out a few bullets and explain the relationship between Scallop and Lobster and what you think is going on?

← PreviousPage 9 of 34Next →