DrScientist

Good leadership makes a huge difference - whether that be big or small. I've seen it myself - in particular it's amazing much of a difference good leadership can make at a huge company.

However at the same time it can also be true that some leaders are either poor or make no difference - but as long as they don't have to react to external change ( or only do so as part of a herd ) they can coast quite happily for years while 'earning' huge rewards.

Leadership is only really tested when the right course is a different one from either the past or your peers.

Perhaps you've watched too many classic StarTrek episodes as a child and are channelling the Spock versus Kirk leadership theme.

All models/data sources are biased - you need to understand inherent biases in any data source.

The first point is a strawman - these models are not going to be used to set foreign policy on Taiwan - it's to write code etc.

Likewise risk of data exfiltration and misuse isn't model specific. Indeed it's not even the biggest data risk - there are much larger risks from the data we know companies like Meta and Google already collect.

Bottom line - a model you can download and run on your own private infrastructure is always going to be safer than anything accessed over the internet - as in that case it's not even just the hosting service that's the issue.

If the online retailer has a no quibble returns policy - then at there is at least a strong incentive to minimise returns, rather than oversell.

> The creation of value is me taking energy from the sun and converting that into a chair.

Which requires money to do.

No it doesn't require money. It requires energy. Money is often involved as a means of exchange, but it's not required.

If what you said is true it would be impossible to live off grid and make a chair.

By focusing on the money you are focusing on the means of exchange rather than the fundamentals - which is energy, knowledge and collaboration.

Let's take a concrete example - the development of Linux - did that require some large VC backed funding models or was it simply people sharing spare cycles for shared benefit? Here the underlying resources required was people's time [1] - not money per se.

[1] Sure some value was exchanged using the medium of money ( Red Hat existed, large companies shared larger cycles by explicitly paying employees etc ) - but that's not lending and interest with money somehow being the time limiting resource. Again don't confuse the means of exchange with the thing being exchanged.

But aren't you confusing the means of exchange with the creation of value.

The creation of value is me taking energy from the sun and converting that into a chair.

You lending me money is you extracting value from artificially being a middleman.

It would have been more efficient to write an IOU to the tool maker, make the chair and pay back the tool maker directly.

Now sure that IOU isn't that fungible - however that highlights one of the absurdities ( if I understand it correctly ) of the current banking system where private banks are able to in effect issue IOU's on their own basis but put mine and your name on it as a guarantor - resulting in the public having to bail out banks when they over extend.

Sure - but that's a meta question of whether the current economic/political systems are delivering good allocation of resources for whatever target function you favour.

And that's complex - even in your simple example you can argue that people have likely improved their capability to dig and fill holes - and the latter skill is particularly valuable in the UK right now.

ie Entropy includes information component :-)

Why not? What am I missing.

If you loan me 10 pounds and I simply keep it or spend it but refuse to pay you back - aren't I the winner and you the loser?

Claiming that the market is a positive sum system, because I use that 10 pounds to invent something that changes world productivity is missing the fact that there is a constant stream of huge energy inputs from the sun - that's ultimately what allows the local entropy to decrease - not that you lent me 10 pounds.

Who funds the majority of cutting edge scientific research?

Is it companies or is it governments?

If governments around the world see LLMs built from public knowledge as pre-competitive as the public knowledge itself, then why wouldn't they sustainbly fund it?

Money isn't lost per se, it's transferred.

ie for every loser there is a winner.

Obvious ones are those who can extract value from this companies now ( high salaries/bonus etc ), swapping worthless paper for real assets, Nvidia and their share holders etc. Investors who hold short positions etc.

This didn't stop the Soviets. They did an enormous amount of very dangerous research with minimal logical application.

Ultimately what stopped them ( and everybody else - let's face it it wasn't just the soviets ), is it's a bad idea.

I think you way over estimate how hard the knowledge bit is in a biological weapon ( if you just want to kill indiscriminately - specific/controlled targeting a whole different ball game ).

The barrier is motivation and kit ( though the kit is much easier to access that say the stuff you need to build a nuke ).

The idea that all that's stopping some disaffected Joe sitting on his sofa at home and suddenly deciding to make a biological weapon is he can't get a set of instructions from ChatGPT is to miss the point.

Ultimatelty it all comes down to are you motivated enough - if you are then the knowledge is out there with or without an AI summary.

The point of taking biometics is they can check if the same person is travelling on the same passport each subsequent visit.

Sometimes they rely on just image matching, and sometimes they ask for an additional finger scan ( but not normally all fingers of all hands like first registration ).

So perhaps if you have a generic looking face, or have changed your appearance since the first visit you are more likely to be fingerprinted.

What you are potentially missing is the public sector wouldn't necessarily be in the same race as the commercial companies.

It doesn't matter if a public effort is a year or two behind the curve ( and hence has dramatically lower costs due to Moores law and the ability to piggy back on research ) - especially as we approach the asymptotic phase of development.

Don't buy the idea that the only thing that's standing in the way of people making bioweapons is the censoring of LLMs.

If you are motivated enough to assemble all the kit you'd need, and actually do it, then you should be motivated enough to find the knowledge to do it without chatgpt etc.

I'd imagine the main thing that's stopping people is biological weapons are a terrible tool to do what most people want to do - which is target specific enemies.

So while the materials you'd need to build such a thing are much more accessible than say nuclear material, it's much less attractive - but if already had somebody mad enough to try - it's already possible.

Indeed.

You could go even further and say that the modern Western preference for whack-a-mole politics rather than long term competence is one of the reasons that countries like China is on the rise.

Sure.

Have you ever listened to music while simultaneously writing a comment on hacker news - and managed the split cognitive load effortlessly?

Sure. Still don't see why the result is at all surprising.

If you couldn't process multiple streams ( I see nothing special about speech - note parsing speech is potentially different from thinking over the top - ie the difference between hearing and listening ) simultaneously it's hard to see how you'd last 5 minutes.

Spending a small amount of time processing stream A, then dropping that and processing stream B for a moment, then swapping back. Just like how a single core CPU can process multiple threads.

The brain isn't a computer with a single CPU - just because computers are built that way doesn't mean the brain is.

Ie it's only surprising if you start with an erroneous computer based model of how the brain works.

The paper is interesting in that it looks at a core interesting issue - which is how conscious attention is managed given we know that behind the scenes it's 'everything all at once'.

If you couldn't process multiple streams ( audio/visual/other senses ) how would you ever be able to monitor the background for danger and context switch?

There is a difference between conscious experience and what's going on in the background.

Sure - though this thinking can lead to the common misconception that CEO's just need to be generically good managers and not experts in the area the company operates in.

I think this is often wrong for two reasons

    - the layers are not entirely disconnected - the devil is often in the details
    - hard to detect BS from subordinates

There are two things here - Steve might well hire smart people to tell him what to do - but in the end Steve made the choices and drove them through.

Management is all about choices with limited information - there is rarely one clear and obvious route forward, but a strategy of following all routes simultaneously will lead to failure as there needs to be focus.

So somebody has to make that choice and then drive that focus through the organisation.

This can be very difficult when the choice isn't clear and obvious - as lots of smart people will have different views.

That's the art of management - you need to make the right choices most of the time. and when you've made the wrong one correct quickly, but you also need to be able to get the organisation to rally behind that choice, even if not everyone agrees.

I think there is a strong cultural effects to whether this is true.

I've worked in multinationals and I've noticed that in the US ( east coast in particular ) it's much more hierarchical - where arguing with your boss in public could be a firing offence.

This is less so in Europe.

I think you are right that the issue is tribal framing.

The question is whether satire will puncture that.

Farage's supporters will be trying as hard as possible to paint Binface as a lefty establishment figure- but satire gives you an oportunity to reframe the debate.

Take the 5 million quid - the defence is it's a gift and who wouldn't take it if offered?

Binface can lean into that - rather than say it's terrible etc - he can say - yes I'd like a bit of that - and ask Farage how he can also get such a donation?

Such a question is in effect asking 'what do I need to do to get a 5 million donation' - which is the same as asking what Nigel did for it.

But it's more difficult to answer with 'none of your business' or I did nothing - implying Binface is equally likely to get it.

But perhaps I'm being over-optimistic.

I do wonder here whether the core problem here is that github is outside your firewall, and so you are always one secret leakage/misconfiguration away from disaster.

this stuff is hard, you will fail a lot, and you will fail a lot more if you don't have an extremely clean environment to do this in

The Oxford Nanopore sequencing technology is one of the most robust to use. You need to buy some kit - but defo doable. Nothing compared pouring your own gel and doing radioactively labelled Sanger reactions :-)

Though you could just go to an sequencing company that services labs ( that just does sequencing outsourcing - rather than a personal genome company ).

Totally agree on the dangers around interpretation.

1,2,3 are dominated by platform stickness or even active lock-in.

Can't say I see the same advantages to stop you switching the model you use.

It seems that enterprises will pay top dollar for service guarantees, integration, and someone they can sue.

Sure. Though it does depend on whether you need regular updates. If you want the model to be aware of the latest research - then fine. However it already does the job, you might prioritize stability over constant change.

It's nobody gets fired for buying IBM all over again.

Except they when they did when IBM was no longer good value for money.

but I don't see any historical analogues

None at all? You mentioned IBM - who is using AIX on IBM hardware in 2026? Who is using Solaris on Sun hardware? It's pretty much all gone to linux on commodity hardware.

Remember Netscape - thew browser company? Killed by Microsoft bundling of IE. How hard would it be for Apple to bundle GLM based services?

With these kind of 'demonstrations', the key question is what have we learnt as a result?

Not sure what that is in this case.

I think you underestimate, in the time before the everyday internet, how many people's world views and values were largely set by their direct personal experience, rather than any media they consumed.

Sure, church and state, and to a lesser extent newspapers all were involved in shaping culture - however the values grounded in personal experience drove many societal improvements in the 20th century.

I don't see that so much now, and the power of persuasion has widened from those big institutions to anybody with money - domestic or foreign and at the same time society is atomising.

I have no idea how.

It's not hard, or a new problem. For example in the UK paid for political campaigning on radio and TV has been banned pretty much since the dawn of those mediums. In terms of print, you can have media pluracy rules, to stop the concentration of power.

Any kind of election spending, at the local and national level, is regulated and capped.

Now people have been using less visible targetted online ads using dark money to try and get around this - in my view this is an extremely serious offence and people should go to prison.

You could go further and make it illegal to spend money on online ads for any kind of political campaign whether it's aligned to a party or not ( ie single issue ) - with penalities for the companies hosting those ad campaigns as well as the people running them.

You could include astroturf botnets as part of this definition.

Sure these things can be worked around - but the point is the risk of being caught breaking the rules, and the political damage as a result, will make people think twice.

One of the problems at the moment is people are getting away with a slap on the wrist - in my view undermining democracy is a serious offence - and it's about creating a culture where that's the prevailing view.

PA was proscibed after the Filton offenses.

A key principle of justice is laws are not retrospective - you can't be put into prison for something that was legal at the time.

Again, yet another example of a failure of natural justice in this case.

Judges don't need jury trials to find facts, they do bench trials without juries at all.

Again you missed the point - trials without juries are for minor things where the person on trial is unlikely to be deprived of their liberty. Juries are there to stop the state from imprisoning people at a whim - indeed they are fully able to choose to refuse to convict, without needing to explain why, even if the facts clearly show the defendents did it.

The safe guard is about 'juries of peers' personal opinions - not about the letter of the law - as the establishment can control that.

Seems like your investment in the outcome is blinding you to the obvious abuse of process.

I can see why the establishment was keen to make examples of these people - however that's not a good reason to undermine the very foundations of justice.