dojomouse

Yup, fair enough - having read the full overview from typesafe I agree they mean “won’t assign a probability to an outcome outside a defined set” rather than “won’t incorrectly assign a probability”.

LLM can be made deterministic too, they’re just generally not because the element of randomness is a feature.

There’s a big difference between deterministic and smooth though. Typical LLMs certainly aren’t reliably smooth, so the small prompt change might product a large and unpredictable output change. I’m not sure if that’s any better with the typesafe approach.

Hallucination is the result of a series of choices of options based on poor predictions of probability (or poor sampling from a probability distribution). As I understand the typesafe approach they’re allowing the set of options to be constrained, but the potential for a misrepresentative probability distribution is still there, and hence so is the potential for “hallucination”.

> The cloud deployment surface covered in our contract would not permit powering fully autonomous weapons, as this would require edge deployment.

… What?? Much of this seems duplicitous, but this isn’t even coherent. Is their implication that it’s not “autonomous” if it involves an api call to an external system? That mere definition would be extremely alarming.

Natural radioactive decay is also an example of fission - just not a fission chain reaction (… except when it is! https://en.m.wikipedia.org/wiki/Oklo )

I think it’s very much a magical fantasy computer :-) A thought experiment, rather than a claim to something that’s been built or exists.

The questions you pose are interesting ones, for the sake of the experiment I think at least:

- Programming

- Loading new software

- Adding new functionality

- Persisting user data

Could all at least in principle be achieved without changing the network architecture, but rather just providing the relevant data at the inputs (eg the full bit stream of the install file of the new software being installed) and having that lead to adjustments in activations (not weights or architecture) across the network which lead to any future inputs to the network resulting in the outputs that would be expected in the presence of the new software. Same deal for the other examples.

As to whether this is practical or achievable at present - not even remotely close in my view. But it’s still an interesting idea, even if just to think about why it wouldnt work and what that implies for future development direction of multimodal networks etc.

Perhaps. I was thinking along the lines of MarkBurns response - ML will allow us to efficiently look in those places we might otherwise only have searched by accident.

If ops point was rather that “accident”/“luck” are uniquely human… I don’t agree. Luck is when probability works out in your favour - and that can happen all the time with any sort of probabilistic search, which is rife in ML.

I think ML is likely to be material to us making many more such discoveries. So much of the current constraint is not in the knowledge to identify the interesting pattern, but the capacity to look for it at scale.

In New Zealand we have a pretty effective system that covers all fossil fuel use nationally, as well as several other greenhouse gas emission classes (notably not agricultural emissions though hopefully they’re included soon). The EU has something similar as do several other jurisdictions. It’s a solved problem at scale.

In NZ we also have an effective system for recognising and incentivising certain classes of forest carbon removals (which I think are a legitimate and important class of credits - unlike avoided emission credits which I agree are junk).

This is extremely cool! Nice work. What data does it rely on and do you have any plans to support other countries?

It wouldn’t help with kinetic energy harvesting from the raindrops as that would go into the funnel as heat.

It might provide a way to harvest the remaining gravitational potential energy of the rain (possible funnel being your roof and guttering) but the only upside is that you could concentrate the energy with something that’s already there (and hence harvest over a smaller area). The amount of energy (and hence value) available would be even lower - unless you had a really high roof.

This is also the reason I abandoned my high school scheme of hydro turbines at the bottom of downpipes.

As the comments below say - you need to be working at the scale of a few major geographic features as a funnel before it starts to get really interesting.

Love you for this! I had exactly the same “solar freaking roadways” thought, although at least that idea qualified by basic theoretical analysis of available energy and area for harvesting and conversion efficiency. It was an obviously terrible idea for other reasons :-) yet it still got a prototype…

I wasn’t sure about the droplet analysis so took your same numbers (25mm/h, 10m/s) and just worked out aggregate mass: 25mm over 1m^2 = 0.025m^3 = 25kg

0.5mv^2 => 1250J/h… so looks like we agree.

And to add a simple economic analysis of why this is such a dead-end idea:

Mawsynram, in India, is apparently the rainiest city in the world with roughly 10,000mm of annual rainfall - 10x the global average.

A given rain energy harvesting panel, deployed there, would generate 500,000J/yr… or 0.138kWh. That’s significantly less than what a typical rooftop 1m2 solar panel would generate in an hour on a sunny day. 0.138kwh is worth around 1.3cents at 10c/kWh.

A big roof might get you $1-$2/year. You couldn’t pay to clean your roof for that. You couldn’t even pay someone to answer an email enquiry about the install costs for your system for that. This solution would have to be VASTLY cheaper than paint to stand a chance of being viable.

There is a reason our existing systems to collect power from rainfall rely on vast existing landscapes and aggregation mechanisms (rivers) to concentrate the rainfall for us.

It is - in my view - a dead idea.

Correct, but what the parent here presents is a theoretical upper bound. A working product wouldn’t even get close. When the theoretical upper bound shows that something could never aspire to more than a vastly inferior alternative to existing proven technologies, the correct approach is to abandon it rather than invest in iterative improvements.

I agree we should keep an open mind regarding creative ways of collecting energy from the environment. But we should also abandon those which are quickly demonstrated to have no meaningful potential even if we were to perfect them.

It also ignores the fact that there’s no need for it to become “alive” or “conscious” to be a threat in the way he describes. It just needs to be an agent with an mis-specified, poorly specified, or maliciously specified goal. And there are already numerous examples of those. The only debate is around capability, and here he makes multiple references to “infinitely” capable. So the whole argument seems like wildly disingenuous strawman, consistent with his attempt to classify all those raising concerns as naive (or corrupt) cultists - not exactly the vibe from the likes of Geoff Hinton / Stuart Russell / Max Tegmark; all of whom generally act with far more integrity (it seems) than Marc Andreessen shows here.

Ironically I think the whole article is motivated by the thing he claims to condemn - namely: he’s a bootlegger, who has an interest in freedom of ai development.

Part 2 is much more interesting. Part 1 was very very weak.

> One reason is that there are some problems that a company will only encounter once it begins testing fully driverless operations. Waymo and Cruise’s problems with fire hoses and caution tape is a good example. A human driver would disengage FSD long before it got into situations like that, which means Tesla would be unlikely to have the training data necessary to train its cars to handle it properly.

The reasoning here seems flawed to me. I assume Tesla use periods when the human is driving for training data, so this is no constraint at all.

No. That wouldn’t be a battery by any meaningful definition, nor have any similarities in implementation or enabling technology or physics.

But it would deliver 24 trillion wh/kg… so by that metric at least we’ve room to progress :-)

Yup - or higher load factors, or electric drive, or higher efficiency engines, or some synthetic fuel produced in a low emission way… all of which the tax provides a nice incentive for :-)

I like the framing ;-) Maybe it could work for climate change deniers: “Tired of THE MAN intruding on your personal freedom to pollute? Regain your independence with solar, electric transport, good insulation, and a low impact diet. Freedommmmm!”

It’s adding a tax to the fuel type for the category, not the category as a whole. The tax will increase the commercial incentive for cleaner shipping which will reduce emissions.

The tax isn’t remotely large enough to result in freight migrating to air. And trucking within the eu is already subject to a carbon tax.

In the case of transferable concert tickets that’s already the ‘solution’. Even if someone ‘poor’ is lucky enough to beat the odds and buy a ticket under the current system at the initial offer price, they still have the same opportunity cost to weigh up - it’s just they’re deciding whether to resell their ticket and realise a gain, or go to the concert and forfeit the gain.

This proposed system would be fairer in that it would at least deliver whatever the maximum amount a rich person is prepared to pay to the artist, rather that some scalper intermediary. Those at the lower end of the willingness-to-pay spectrum are no worse off, except to the extent that you think they’re disproportionately lucky, disproportionality motivated (which I accept is possible), or are themselves looking-to/willing-to scalp tickets.

It also seems however that it’d be pretty trivial to just prohibit resale of tickets and require ID at the venue. Artists/promoters might make less money but it’d preserve equality of access and largely eliminate scalping.

I have an innate desire to write a book. But I haven’t because the effort would be significant and I have other things to do, including work. I am glad that the many professional authors whose work I’ve enjoyed had financial reward providing them both the motive and means to write for me.

You could argue I have an innate desire to care for people inasmuch as I was a volunteer ambulance officer for a while. But I’m pretty glad we pay doctors and nurses and paramedics to dedicate their working lives to doing an excellent job of that stuff rather than just assume that pro bono efforts will see us through. I think it’s enormously naive to assume we’d lose nothing if we took away art as a profession - especially since it’s so obvious it would be a dumb idea to do away with many other professions.

Also I find it hard enough finding books and music I genuinely love even WITH the profit motive at work and giving people the ability to dedicate their lives to it!

Do what Canada does; add a carbon tax and distribute the proceeds. This is generally a progressive tax scheme so low income people end up with MORE money.

I’d argue another consideration is the land occupied by the corn - it’s the real constrained resource, not the corn growing on it, and were the demand for corn displaced by something else (like direct air co2 capture) then that same land could sequester carbon far more effectively with e.g. trees.

Was going to ask the same question and scrolled through the comments to happily find yours. The answers basically seem to explain why there's no robotaxi business model NOW though... I still don't buy the argument that there wouldn't be one with suitable (relatively low marginal cost, eg Tesla system) self driving tech. Yes, uber drivers work for less than min wage, but that's still expensive compared to cost of capital for a (speculative commercial) self drive unit as an add on.

There were numerous direct references to Valve, explanations of issues in the context of Valve, and commentary from Valve employees.

As someone who really liked the ideals and concepts of the Valve handbook but struggled to see how to extend them to my own (hierarchical) workplace, I found it quite enlightening.

Haven't looked at the proposal, and not commenting either way on the risk of using water in this context, but the water/reactive-metal issues have been widely investigated in other liquid metal reactor designs and generally intermediate cooling loops with some benign medium are used.

Edit to expand; having now looked it seems some of the intermediate loop designs have the SAME material (e.g. sodium). This at least means the loop at risk of mixing with water is isolated from the core (where an explosion is difficult to manage). I thought I'd seen other designs, just can't find them now.

I suspect we're going to end up seeing evidence that the first accident was also the result of known risks being down-played / swept under the rug in the name of profit. One perspective is that the first and second accidents happened because the band-aid fell off, and the first accident should have been the cue to use stickier bandaids. But it seems very unlikely that no engineer anywhere through the development said "This design is for shit and is WAY too dependent on a bandaid that can't realistically be adequate".

I should qualify that a bit...

- The promoters are likely to present at least a realistic (if not optimistic) estimate of mid term cost, the technologies involved are all mature (so the potential for further reduction I'd low), and the estimated cost is not materially better than battery energy storage (worse, in fact, than near term forecasts, and much worse than midterm potential battery energy storage cost).

- The siting flexibility is poor, which means grid connection will be slow/expensive.

- The power density will be pretty bad (which wouldn't matter for bulk energy storage... but it's too expensive to compete in that market).

Maybe it could become more attractive with much taller cranes?

If the promoters themselves are only claiming $150/kwh as a goal this doesn't have a chance of success given its other limitations.

Gah, what a formatting train wreck :-/ sorry...

I looked into this in a lot of detail about 5 years ago. There were a couple of studies, and one somewhat-active company I saw (Stratosolar - didn't seem to be very well resourced/capable).

My main focus was on aerodynamic modelling and panel positioning methods for various structure sizes, and the resulting LCOE. Main issues I found were: - Weather conditions in the stratosphere aren't well understood; most of the time pretty benign, but there are a bunch of extremes which could have a significant impact on the structural requirements. - It's basically a tradeoff of panel-cost/conventional-installation-cost vs aerostat-cost/non-conventional-installation-cost. The aerostat is definitely not going to be cheap, so having your panels on an aerostat has to result in a bunch more energy per PV-element than having them on the ground. - Having the aerostat option come out on top gets more difficult as PV gets cheaper. Let's say you get 2x energy from PV on an aerostat vs installed on the ground. That means the aerostat option will be competitive with the ground option as long as the total installed cost (per watt) is less than 2x the terrestrial installed cost. If the terrestrial installed cost reduces by a factor of two (and it's reduced by more than that since I did the analysis!), you suddenly have to reduce the marginal cost of your aerostat option by 50% just to remain competitive! - To be economic and sufficiently robust to expected weather, these structures have to be enormous; the architecture that seemed most promising to me (from memory) was cylinders of length 4km and diameter 1km (roughly 1GW electrical output peak, more like 500-600MW annualised). They're at least semi opaque, and are tethered around 20km altitude (and can drift within a ~10km radius around the tether point). At that altitude they're visible from several hundred kilometers away, and they look huge - 15x the width and length of the largest cruise ships. - It doesn't help THAT much with seasonal variation away from the equator. Summer output in northern europe is still 2-3x winter output, so you need long term storage or an energy dump.

So... I think it's super interesting, but I don't think it'll ever be commercially attractive vs either terrestrial installations, or space. The main nice thing is that it's still pretty easy to get the power back down to earth with high efficiency... in contrast to orbital solar.