fpoling

There is a difference between SoaceX and Anthropic. The former is objectively ahead of competitors by many years while Anthropic is at best by few months and pressure from open weights models is increasing.

So even if one believes in AI the rational choice is to invest into hardware that AI companies needs (GPUs, gas turbines, batteries, solar panels)

The article suggests exactly that requiring to delay IPO until 2 consequent quarters of detailed financial audits will be available from Anthropic. Without that there is simply not enough information for an investor to make a rational choice.

And how much money the shop looses from the attack? It sounds it wiped out all the saving remote automation brought.

I should have mentioned language/framework for JS. The thing with Elm is that it gives a straightforward way (even if verbose) to design the whole app around persistent data structures. That gives so many benefits especially long term (time-travelling debugging support, easy insights into app state, full coverage by unit tests etc.). On the other hand my experience with React is that the state hidden in all those functional hooks is hard to debug and reason about with completely unpredictable and non-intuitive behaviour. I think the switch from class components to hooks was a big mistake.

An important trait of good functional code is use of persistent data structures. JS code using React is anything but that with state hidden across numerous tiny functional closures.

The only language/framework that is truly functional in that sense is Elm.

This notion of instantly getting knowledge hides the weirdness of QM. It is not surprising that when one reads a note from a box one instantly can get the knowledge about events in a distant galaxy. What is really weird with Copenhagen interpretation is that content of the note about what happens in a distant galaxy is not determined until the observer looks at it.

Transactional interpretation of quantum mechanics is compatible with relativity and it is it that deserves more attention than Bohm interpretation. Unfortunately it uses a terminology like retrocasuation which sounds even stranger than how Copenhagen interpretation explains QM. So as long as experimental predictions stays the same I see little chances that transactional or similar interpretations will offset Copenhagen mainstream.

For ransoms bitcoins are still transactional. There are sort-of black market exchanges that can exchange bitcoins from blacklisted wallets into other assets or new bitcoins by taking a big cut like 30% percent or more. But with those there is always a risk that the operation would be discovered and all new wallets will be blacklisted. For transactional operations such risk is fine, but for store of value it is very bad.

Bohm interpretation is not compatible with relativity. One can argue that Copenhagen interpretation is also not compatible, but then from practical purposes both interpretations are equivalent and predict the same results so why bother with changing explanations when the new one still does not address the fundamental problem.

I doubt Escobar would use Bitcoins to store funds. They are traceable and are too volatile to store illicit gains. But to carry transaction he would use them on massive scale surely. Cash is sort of the opposite. Not very convenient to carry big illicit transactions but as a store of value it is a way better thing.

GC in Java is the best in world with a lot options and knobs that allows to tune to particular application requirements. Still some trading platforms using it disable GC and just ensure enough memory is available so memory leaks during the trading hours would not matter before the application will be restarted.

I worked on a project with a network server written in Go with soft realtime requirements. In theory Go should be very good fit for it especially given GC tuning towards low latency, not throughoutput. In practice the latency spikes from GC were bad and there were no good options to resolve those given the deployment constrains and the amount of third-party libraries used.

Weak measurement experiments have shown that one can observe the whole path of a particle in a double-slit experiment and there is nothing wave-like in the results.

The experiment does not demonstrate wave-like properties. It can be interpreted that way, but purely from experimental properties one can also interpret the result that photon traveled by curved path which brings path-integral formulation of quantum mechanics when there is nothing wave-like but there are just probabilities of a particular path.

There is no interference pattern from a single photon as it hits the registration device at a particular place. But collect a big enough sequence of single photon measurements and it emerges.

This is the reason for statistical interpretation of QM that states that a wave function is just a way to calculate probability with no inherent meaning.

So wave-particle duality is just an artifact of Copenhagen interpretation of QM, not something that one can observe. What one clearly observes that photons travel through a region where they should not classically be and never appear in the region where they should be classically.

Note that even classically waves does not exist as they are just an artifact of motion of collection of molecules that on average can be described by wave-like equations.

Germany has not learned anything from the war in Ukraine and still think that dependence on gas imports from other countries is fine even for backup.

The main advantage of gas plants is that they support variability. In a world with enough battery storage to last for days that would be irrelevant. And as an insurance against rare events it is better to keep coal plants as transportation and storage of coal is vastly more resilient to interruptions.

Keeping coal plants around is cheap to cover that. Longer term things like producing ammonia during summer and converting it to electricity in fuel cells will allow to remove needs for fossil fuels completely.

Just by scaling up the existing technologies the full cost including installation and maintenance of a grid-scale battery in 5 years will be below 100 usd/kwh. So the price is max 400 billions or 1% of US GDP.

Batteries already kill peaker plants as a grid-scale battery that lasts 4 hours became cheaper than running natural gas turbine to cover the evening demand peak.

And vise versa, a proof that P!=NP do not exclude an algorithm that is fast enough for practical n.

Particles do exist for practical purposes. If one can distinguish individual photons, then the picture is nothing but a bunch of particles with only averages producing wave-like pictures. But the puzzle is that they do not travel in straight lines and we cannot predict individual point where they will hit.

I have used gandi.net for many years. They provided good support and offered things like free email forwarding. Then they were bought and the new owners just double or more prices and made anything extra including email a ridiculously expensive extra.

Cloudflare has verified bot program.

In 10 years when the batteries will cover 24 hours of demand to cover few weeks without sun and wind nuclear will be too expensive as reactors requires maintenance even when they do not work. Coal will be vastly cheaper beating even natural gas as one can store coal in bulk.

But right now we do not have excessive capacity and smoothing the curve for nuclear allows to burn less coal and natural gas. Essentially right now batteries are killing peaker natural gas plants, not nuclear or coal.

Storage is still expensive. But if there is enough storage to cover the demand for couple hours then it can also be used to store excessive energy from nuclear making throttling down the reactor output much smoother. In few years when the storage will be cheap enough to last the whole night it will kill nuclear, but while the capacity is low why not to use it to make running nuclear more efficient.

Modern coal plants are more efficient and have bigger daily throttling capacity offsetting some of the above.

EDIT: batteries also help on China scale to throttle coal output much more smoothly and efficiently. Eventually the batteries will kill coal but for the next 5 years in China they are making coal plants cheaper to operate.

The problem with nuclear is that their daily throttling capacity is low leading to those minimal price contracts. So it makes sense to pair nuclear with batteries as coal plants in US started to do. With that having nuclear as a base load during winter or prolonged bad weather would still make sense.

Batteries are still too expensive to cover the whole night. So it will take few years before they will threaten nuclear base load. But they are already cheaper to serve electricity from solar or wind during 4 hours evening peak consumption than running natural gas peaker plants. Plus they reduce the scale of grid upgrades in many cases.

Battery at a coal plant also allows it to operate at less variable output which is more efficient. Modern plants can throttle more than the old ones, but even with them a battery to smooth the load curve make sense economically to use less coal.