HNHacker News
TopNewBestAskShowJobs

sgt101

7,940 karma · joined January 24, 2012

nothing to see here, move along, move along.
submissionscomments
sgt101··on ARIA, the UK's Bet to Build Scientific Revolutions
>He said that China and the world don't see the UK as a big player anymore. Which is true.

So what? You can see people however you want, but you still have to live with them in your world and they have a vote on what happens next. This is what Russia discovered in Ukraine, and the USA in Iraq before it. China may want to boss everyone in the world, but it can't - it can't have everything it's own way and run the world as a Marxist Leninist 1984 style nightmare, we won't allow it.

>UK has it's legacy advantages, UN permanent seat, nuclear weapons etc but the empire is long gone and it grows weaker by the year.

Well after WW2 we had 270% GDP to debt ratio and the country was bombed into bits. British industry had retooled for a war economy and there was no capital to retool meaning that the fundamental disadvantages for industries like shipbuilding where British yards were located in the wrong places on the wrong rivers became sharply plagent. I grew up in the 1970's and there were still bomb sites in London and Portsmouth then. If that was a legacy then you can keep it.

>After the war, Germany was at its lowest point and the UK squandered it's lead and fell below Germany.

I agree, we should have kept Germany as an agrarian low population country so as to dominate continental europe... oh no, wait: I just realised! That would be both evil and stupid! Instead we should have Germany develop into a liberal democracy full of enlightened well educated Germans and make sure to be good friends with them so that we can work together in the world to shape it for the common good.

>All of Europe + USA failed to bring Russia to it's knees no matter what they tried.

Or we were trying the "Germany mark 2" move, but the Russians are so stupid that they invaded Ukraine and manged to ruin themselves for the rest of the 21st century? Time will tell. Perhaps ask the guys with no hands and faces begging on the Moscow underground what they think?

>The world is changing and in the new world order UK is slipping down the ladder.

For sure, but who cares? No one serious in the UK wants to be in charge of the world and frankly that's been the case since about 1919.

What we want is to sit on our little island with our nice wind turbines and good internet and make interesting things while enjoying our gardens. I think that fundamentally that's what 99% of the worlds population would like to do but there's 1% playing "lets kill the XYZ" or "I am due the respect and tribute of the people of YZX" as if this was 1850 and no one has nuclear weapons.

sgt101··on ARIA, the UK's Bet to Build Scientific Revolutions
Well, a Marxist Leninist analyses the global scene and concludes that communism is inevitable.

We've been here before.

sgt101··on Tell HN: Beware confidentiality agreements that act as lifetime non competes
Has anyone been caught by this? As in sued or prevented from working?
sgt101··on Chemical knowledge and reasoning of large language models vs. chemist expertise
Also, books, books are really good for finding knowledge !

Seriously LLM's as a cultural technology cast them as a super interactive indexing system. I find that's a useful lens to use to understand this kind of study.

sgt101··on Childhood leukemia: how a deadly cancer became treatable
Is there any plausible connection between these statistics and the ending of atmospheric bomb testing? Also the various catastrophic releases of plutonium into the atmosphere from satellite reentry?

Could the gradual reduction be attributed to the decay of radioactive isotopes in the environment?

sgt101··on The Gentle Singularity
>This is a deliberate choice on the part of the model makers, because a fixed checkpoint is useful for a product. They could just keep the training mechanism going, but that's like writing code without version control.

Training more and learning online are really different processes. In the case of large models I can't see how it would be practical to have the model learn as it was used because it's shared by everyone.

>To the extent I agree with this, I think it conflicts with your own point about us not knowing how human minds work. Do I, myself, have deeper truths? Or am myself I making surface level association after surface level association, but have enough levels to make it seem deep? I do not know how many grains make the heap.

I can't speak for your cognition or subjective experience, but I do have both fundamental grounding experiences (like the time I hit my hand with an axe, the taste of good beer, sun on my face) and I have used trial and error to develop causative models of how these come to be. I have become good at anticipating which trials are too costly and have found ways to fill in the gaps where experience could hurt me further. Large models have none of these features or capabilities.

Of course I may be deceived by my cognition into believing that deeper processes exist that are illusory because that serves as a short cut to "fitter" behaviour and evolution has exploited this. But it seems unlikely to me.

sgt101··on The Gentle Singularity
>The diversity is itself indicative, though, that intelligence isn't bound to the particularities of the human nervous system. Across different animal species, nervous systems show a radical diversity. Different architectures; different or reversed neurotransmitters; entirely different neural cell biologies. It's quite possible that "neurons" evolved twice, independently. There's nothing magic about the human brain.

I agree - for example Octopus's are clearly somewhat intelligent, maybe very intelligent, and they have a very different brain architecture. Bees have a form of collective intelligence that seems to be emergent from many brains working together. Human cognition could arguably be identified as having a socially emergent component as well.

>Most of your critique is surface level: you can add all kinds of different structural diversity to an ML model and still find learning. Transformers themselves are formally equivalent to "fast weights" (suppression and control signals). Continuous learning is an entire field of study in ML. Or, for injury, you can randomly mask out half the weights of a model, still get reasonable performance, and retrain the unmasked weights to recover much of your loss.

I think we can only reasonably talk about the technology as it exists. I agree that there is no justifiable reason (that I know of) to claim that biology is unique as a substrate for intelligence or agency or consciousness or cognition or minds in general. But the history of AI is littered with stories of communities believing that a few minor problems just needed to be tidied up before everything works.

sgt101··on The Gentle Singularity
>which is in many ways similar to our subconscious thinking

this is just made up.

- we don't have any useful insight on human subconscious thinking. - we don't have any useful insight on the structures that support human subconscious thinking. - the mechanisms that support human cognition that we do know about are radically different from the mechanisms that current models use. For example we know that biological neurons & synapses are structurally diverse, we know that suppression and control signals are used to change the behaviour of the networks , we know that chemical control layers (hormones) transform the state of the system.

We also know that biological neural systems continuously learn and adapt, for example in the face of injury. Large models just don't do these things.

Also this thing about deeper and deeper realities? C'mon, it's surface level association all the way down!

sgt101··on AI Lied to Sky News Journalist (UK)
A very useful demonstration of how non technical people interact with AI and how that interaction can collapse catastrophically.
sgt101··on A deep dive into self-improving AI and the Darwin-Gödel Machine
Super interesting.

I will have to read it - I will be looking to figure out if the tasks that they are working on significant/realistic? And are the improvements that they are finding robust?

sgt101··on A deep dive into self-improving AI and the Darwin-Gödel Machine
I was trying to pick n-shot examples from a data set. The idea was that given 1000s of examples for a prompt finding a combination of n that was optimal could be advantageous, but for n's that are large then bruteforcing the combincation would be impossible... so can we find an optimal set with an efficient search?

But the problem was that the search space wasn't informative. The best 1 example didn't feature in the best 2 examples. So I couldn't optimise for 5, 6,7 examples..

sgt101··on The Sky's the limit: AI automation on Mac
My brother, you haven't drunk the right coolaid.

It's not impressive because text is a shit interface for interacting with a machine. GUI's are much better and we established this in like 1982.

Also what they are doing in that demo is inserting a machine shim in an interaction between people, which is what people, time and again, say they do not want.

sgt101··on A deep dive into self-improving AI and the Darwin-Gödel Machine
I spent a lot of time last summer trying to get prompts to optimise using various techniques and I found that the search space was just too big to make real progress. Sure - I found a few little improvements in various iterations, but actual optimisation, not so much.

So I am pretty skeptical of using such unsophisticated methods to create or improve such sophisticated artifacts.

sgt101··on Everything We Learned About Lockheed's "QuadStar" Stinger Missile Replacement
if the enemy can't actually prosecute the offensive successfully. If they can, you find yourself kneeling in front of a hastily dug trench.

You are right about the US and Russia though. It has suited everyone apart from the Russians and Ukrainians for the Russians to pick up this tar baby. I don't know if the rest of the world sees it, but a big part of the mess is courtesy of the British as a part of an extended and excruciating revenge play for the Russians messing about in Salisbury. Vashe zdorov'ye you bastards.

sgt101··on Everything We Learned About Lockheed's "QuadStar" Stinger Missile Replacement
>"Expensive (or even relatively cheap!) manpads make no sense vs drones economically, and manpads vs low flying/nap of the earth helicopters, especially in forested terrain have always been of limited effectiveness."

My point is that this isn't just an competition of economic exchange. The capability that a viable organic air defence option gives to a military unit shouldn't be understood in terms of "we spent $50m while they spent $500k". This is because if the $50m spend enables your units to successfully prosecute an offensive the value is potentially 1000x, or more.

sgt101··on Everything We Learned About Lockheed's "QuadStar" Stinger Missile Replacement
100% - this must change.
sgt101··on Everything We Learned About Lockheed's "QuadStar" Stinger Missile Replacement
I don't think we know, but it does look like the answer is Yes.

Both Iraq in 1991 and Iran in 2024 had extensive, developed, and expensive AA setups. The USA destroyed the Iraq assets by attrition, Israel surgically removed the Iranian ones.

Russia apparently has no SEAD or DEAD capability because they didn't and haven't even attempted to remove the Ukrainian air defence network.

I think this shows one of the key requirements for future European Air Force structures. Long range strike and SEAD/DEAD are going to be fundamental for creating the conditions for military dominance in the east.

sgt101··on Everything We Learned About Lockheed's "QuadStar" Stinger Missile Replacement
yes - agree, people look at a picture and don't look at what's not in the frame or what other pictures can look at.
sgt101··on Everything We Learned About Lockheed's "QuadStar" Stinger Missile Replacement
Sadly this is a past tense thing now since Rukban in Jan 2024 where 3 U.S. soldiers unfortunately were killed by an Islamist (Iranian backed) group with a drone.
sgt101··on Everything We Learned About Lockheed's "QuadStar" Stinger Missile Replacement
If you send your troops or vehicles forward without defense they will all get killed. You are not going forward in order to exercise your vehicles and you are not defending your vehicles and troops for the sake of simply preserving them. You are doing it in order to be able to go forward and win.

Armies that can't conduct an offense are unable to win wars. At best they are able to wait for wars to finish. Remember the nightmare of the columns in the north of ukraine in 2022 - TB2s (remember them?) ranging up and down blowing up operationally relevant assets without being hindered. If those troops had had a way of protecting themselves it's possible that they could have been resupplied and evacuated. If that had happened, or if they had been able to refuel and push south Russia might have won, it was pretty close anyway. Imagine the value of avoiding 3 years of this conflict.

sgt101··on Everything We Learned About Lockheed's "QuadStar" Stinger Missile Replacement
I think that they give adversaries another card to play.

Drones mean that you can do a bunch of tasks that previously needed air dominance without that dominance, or even when you are dominated - like the Houthis are showing. With drones you can surveil and strike even while your enemy kills everything that goes 20m above the ground.

If you can contest the air at the same time I would say that their role becomes greater - as the Ukrainians have showed. But drones do make air dominance less decisive, even as it remains very desirable.

sgt101··on Everything We Learned About Lockheed's "QuadStar" Stinger Missile Replacement
Ummm I don't remember that - can you give some more details?

I do remember them running articles like "what should NATO do if Russia invades". Which isn't a dismissal.

Of course, it would have been a really bad thing to dismiss & deride a Russian invasion that took Kiev in 4 days and ended with the partition and neutering of Ukraine with less than 1000 Russian dead. But I don't live in that reality, I live in the reality in which the Russians managed to pull of an even bigger foreign policy mistake than Iraq 2003. It's not got to the all time high of June 1941, but it's getting closer with every week that goes by.

sgt101··on What If We Had Bigger Brains? Imagining Minds Beyond Ours
"The fact that your quote says "Mental processes are not, or not only, computational processes." is the icing on the cake. Consider the unnecessary wording: if a process is not only computational, it is not computational in its entirety. It is totally superfluous. And the assumption that mental processes are not computational places it outside the realm of understanding and falsification."

Let's take this step by step.

First, how adroit or gauche the wording of the quote is doesn't have any bearing on the quality of the concept, merely the quality of the expression of the concept by the person who formulated it. This isn't bible class, it's not the word of God, it's the word of an old person who wrote that entry in the Stanford encyclopedia.

Let's then consider the wording. Yes, a process that is not entirely computational would not be computation. However, the brain clearly can do computations. We know this because we can do them. So some of the processes are computational. However, the argument is that there are processes that are not computational, which exist as a separate class of activities in the brain.

Now, we do know of some processes in mathematics that are non-computable, the one I understand (I think) quite well is the halting problem. Now, you might argue that I just don't or can't understand that, and I would have to accept that you might have a point - humiliating as that is. However, it seems to me that the journey of mathematics from Hilbert via Turing and Godel shows that some humans can understand and falsify these concepts.

But I agree, Wolfram is not under any obligations to consider embodied congition, thinking around enhanced brains only is quite reasonable.

sgt101··on Trading with Claude, and writing your own MCP server
I remember that John McCarthy used to condemn AI research as being a series of "look mom no hands" demos, but now we've got to "look mom I face planted". This work would be valid if it did things like :

- Describe an intelligent effect rather than "invoked an API" - so something like the system evaluating the portfolio and the market and deciding what should be sold or offering buying opportunities based on some inference (this is not a great example for many reasons).

- Measure and report the systems performance. The write-up says that the LLM fails, ok... how often?

- Describe the failure cases, provide some theory as to why some things succeeded and others didn't.

- Provide a way forward. What's next?

Without doing these things this work isn't helpful and is part of the AI/Agent/MCP hype. Basically it's Bored Apes for AI.

sgt101··on Trading with Claude, and writing your own MCP server
Yeah - feels like lunacy to me. I love the way that they get to the end of the article and then write that it doesn't really work!
sgt101··on Denmark to raise retirement age to 70
You can find the info here: https://www.ons.gov.uk/peoplepopulationandcommunity/birthsde...

There was a steady rise in life expectancy at retirement age in the UK until 2020, then something happened and it's fallen significantly. My guess is that it will rise rapidly over the next 15 years and this will be regarded as some sort of triumph for the public health system (unless there is another pandemic of course).

sgt101··on Denmark to raise retirement age to 70
Higher life expectancy more like, the state can't afford for people to spend 20 years unproductive. This is a much less acute issue in the USA and Russia where health is poor and life expectancy low.
sgt101··on AI in my plasma physics research didn’t go the way I expected
There is a lot of retrieval in the behaviours of LLM's, but I find it hard to characterize it as overfitted. For example, ask ChatGPT to respond to your questions with grammatically incorrect answers.
sgt101··on I got fooled by AI-for-science hype–here's what it taught me
Apart from self driving and autonomous flying...
sgt101··on AI in my plasma physics research didn’t go the way I expected
I think that while the mathematics of neural networks are clearly completely understood we do not really understand why neural networks behave the way that they do when combined with large amounts of real world data.

In particular the ability of auto regressive transformer based networks to produce sequences speech while being immutable still shocks me whenever I think about it. Of course, this says as much about what we think of ourselves and other humans as it does about the matrices. I also think that the weather forcasting networks are quite shocking, the compression that they have achieved in modeling the physical system that produces weather is frankly.... wrong... but it obviously does actually work.

← PreviousPage 8 of 34Next →