Knives are out for the Dr
I wondered when this would come. It seems likely Labor loses the state election in Victoria in five weeks’ time. Not a slam dunk but likely (coalition are $1.40). It’s expected though, so won’t necessarily cause a national shock wave and probably won’t alarm Albanese that much. Then we have NSW in March next year. Labor are in front there on a two-party basis (all polls are listed here) but their overall vote will collapse.
If I was Albanese, I’d leave Jim Chalmers where he is. Watch as Labor gets hammered in Victoria, then see how it's going into next March. If it’s looking bad, Jim will be shuffled off to some minor ministry, where we won’t have to hear ‘a good thing’ quite so often. Albanese isn’t the brightest, but he is a much better politician than Jim Chalmers and he’ll let him suffer right up to the moment that the net impact on the PM himself is negative. That’s a while away yet.
The final straw was the Intergenerational Report Jim delivered a few weeks back on just how great everything really is. In that report, he gleefully shared this chart about how, for the next 40 years, Australian governments will spend more money than they earn in taxes. You couldn’t run a business or a household with that kind of budgeting. It's just completely ridiculous not to aim for better. The sad part is, this being a government written report, it is clearly an optimistic appraisal. If everything goes well, we still spend more than we have. More than likely, the outcome will be worse.
Then there is the energy chart. This one is a complete fantasy. Wind is going to be ⅓ of the energy mix? Yet today Australia does not have a single working offshore wind turbine. Geographically the country is unsuited for it because of the distances involved. Solar yes, wind just won’t work. Look at the chart though, if we get a cloudy week with no wind, 80% of our energy will vanish and the entire economy will collapse.
More importantly, the annual electricity generation rises from 220 TWh today to about 450 TWh. It’s a growth rate of about 3% per annum which would be good. Except….the population of Australia is, according to the report, rising from 27 million today to 36 million in 2050.
So (if wind delivers) per head of population you are looking at about 1.8% growth per annum. Energy per capita being, at least in my view, a far better proxy for national growth and wealth than any other measure. All the while, the government forecasts its spending at 2% of GDP in excess of what it plans on collecting.
Check out this word salad of fantasy.
Which local inventions? Which foundational models? Well there is one. Kangaroo LLM no less! A project that launched in 2024 and continues to ‘compete internationally’ even though it has no model, no website, no GPUs, and apparently, no progress.
Overall then, Jim’s Intergenerational Report was horrible. Lacked any ambition, lacked any vision and lacked any recognition of the advantages that Australia clearly does have (mining for example qualifying for only 14 mentions). Accordingly, yes, I think the Prime Minister will eventually sack Jim Chalmers. He’ll wait for the moment of greatest personal advantage to do so.
In the end, the results don’t lie. He has been a terrible treasurer.
It begins
So DeepSeek is now banned in US government organisations. Hard to object to that, it’s sensible and it will surely disturb the reach of the Chinese models. It was a big week for DeepSeek too, releasing their new, excellent and cheap model.
Snappily known as DeepSeek-v4.1-Flash-1017, it competes with the top frontier models on many of the benchmarks. That’s not really the point though, it’s the cost. In this chart you can see cost as a log scale and the DeepSeek model is completing the task for 1/50 the cost of GPT Astra and Claude.
One way to slow down the advance of Chinese AI is to stop using it because they certainly collect and train off the data (just like the US does). It is an exceptional performance though and I’m glad DeepSeek are doing it because it's really pushing the Americans hard (and the team at Kangaroo LLM).
As a matter of interest I think it worthwhile downloading these models personally and storing them at home. I think they will be banned in Australia at some point soon. They are large so you need a decent hard drive. They are open-source and their creators actively encourage people to download and use them. So, I recommend:
Today, unless you have a very large sophisticated computer, you won’t be able to use them. In 18 months it will be different and at the speed the world is moving, you might be glad you have those files.
JP Morgan
So having hated Bitcoin forever, JP Morgan now thinks it will be the best performing asset class of the next decade? To be fair, in the body of the report there are heavy caveats, as well as the 300 pages of “this is not financial advice”.
What do I think? No, I don’t think it will be the best one, I think it will do very well because the macro-chaos now upon us (which I think is overwhelmingly positive macro-chaos) will help bitcoin a lot. Lots of things will change though, there are now all sorts of niche possibilities that will become real very quickly.
If I had to bet on a sector (just for interest) that would beat bitcoin it will be health. The strides in health and cures might be enormous in the next 10 years. Note OpenAI just bought a drug development business this week because they know their tech can massively accelerate what the company is doing.
So JP Morgan will be wrong, but not that wrong.
Kraken
I’m surprised it took this long. Banks are far too worried about their reputations to buy cryptocurrency exchanges, it was always going to be the other way around. So it is, Kraken’s parent has agreed to buy Saldo Bank. Hardly Bank of America, Lithuania based, online only but it does come with a European banking licence and that is valuable. Not least because the risk of being debanked is very much lower when you are in fact the bank.
It’s an interesting environment for banks right now. Five out of six of the “Top Companies” in Australia are banks. How? What innovation have they shown? What have they created? I’m not saying they aren’t well run because a couple of them really are, the others not so much, but really?
Japan’s stock exchange was like this in the late 1980s. The United Kingdom was like it in the mid 2000’s. Iceland managed it and Ireland too. It was always at the peak of a housing boom. The same thing happened every time. If your banks are the best thing you have, it doesn’t actually stack up. Logically, I don’t see how they can outperform, or be worth orders of magnitude more than the things they are supposed to be supporting.
Euro-Trash
It wasn’t good, France. Limoges Town Hall burned down, not to mention the oldest food market in Paris, which also lies in ashes. Yet, the French continue to be French. With fires burning and hooded youths torching the whole city, this couple just carried on amidst the flames and the rioting. Smoked and drank their Cognac.
It’s a real relief to know that some aspects of culture aren’t flammable. Vive la France.
