As was pointed out to me on Twitter recently Betteridge’s law of headlines states that “Any headline that ends in a question mark can be answered by the word no.” That’s not entirely correct, I think. In many cases, the answer is “wrong question”. At any rate, that’s how I treat it in my recent Inside Story review article, where the headline is followed by the standfirst text “Three new books challenge lazy thinking about job-stealing robots and infallible algorithms”. I talked about Ellen Broad and algorithms last time. Now for Tim Dunlop and robots.
This is an extract from my recent review article in Inside Story, focusing on Ellen Broad’s Made by Humans
For the last thousand years or so, an algorithm (derived from the name of
an Arab a Persian mathematician, al-Khwarizmi) has had a pretty clear meaning — namely, it is a well-defined formal procedure for deriving a verifiable solution to a mathematical problem. The standard example, Euclid’s algorithm for finding the greatest common divisor of two numbers, goes back to 300 BCE. There are algorithms for sorting lists, for maximising the value of a function, and so on.
As their long history indicates, algorithms can be applied by humans. But humans can only handle algorithmic processes up to a certain scale. The invention of computers made human limits irrelevant; indeed, the mechanical nature of the task made solving algorithms an ideal task for computers. On the other hand, the hope of many early AI researchers that computers would be able to develop and improve their own algorithms has so far proved almost entirely illusory.
Why, then, are we suddenly hearing so much about “AI algorithms”? The answer is that the meaning of the term “algorithm” has changed.
Read More »
I’ve just done a review article for Inside Story. The headline is Will a robot take my job? but the central point is that this is the wrong question to ask. While technology has a logic of its own, what really matters is our current set of economic and social structures, the financialised version of capitalism commonly called “neoliberalism“.
Read my review and buy the books!
In my recent piece in The Guardian, mostly about Adani, I observed
The paradoxes of Adani are mirrored in the global coal market. Despite a small increase in 2017, global coal production is below its 2013 peak. Yet prices have recovered strongly, yielding big profits to existing miners and offering a seemingly tempting prospect for new mines.
It turns out that this isn’t quite right. The benchmark Newcastle price, for low-ash coal with a heat content of 6000kcal/kg has risen strongly, to the great benefit of companies like Yancoal, Glencore and Whitehaven. It turns out, however, that this increase isn’t representative of the broader market. Prices for lower quality coal with lower heat content and higher ash content haven’t moved at all, with the result that the premium between higher and lower grades has grown dramatically.
What’s going on here? One possible explanation is that Yancoal and Glencore, who produce the majority of Australia’s high-grade coal, have engaged in successful cartel behavior. Another is that the premium reflects shifts in demand (with China and India increasingly rejecting high ash coal, while Japan continues to demand high grade coal) and supply (few new mines are opening, and this has a bigger effect on the smaller market for high grade coal).
Whatever the explanation, most analysts agree that it is more likely to be resolved by a decline in the price of high-grade coal rather than an increase in the price of low-grade coal.
Where does Adani fit into all this. Most of the discussion I’ve found focuses on the premium between 6000kcal/kg and 5500 kcal/kg. Coal extracted from the Carmichael mine would be much lower quality, below 5000 kcal/kg.
That’s the title of my latest piece in The Guardian, responding to a Matt Canavan spray against critics of a recent CSIRO report canvassing options for expanded irrigation in Northern Australia. Interestingly, although Canavan comes across as a typical North Queensland developmentalist (for whom I would have some sympathy) he’s actually from the South-East corner, a UQ economics graduate and a former senior official of the Productivity Commission. Ten years ago, he’d have been debunking CSIRO in exactly the way I do in my report.
After my piece came out, there was a bit of a kerfuffle on Twitter over whether CSIRO had really proposed a dam on the Fitzroy. Their report didn’t do any new analysis of major dams (a point they stressed) but dusted off a couple of existing proposals, then did a more detailed analysis of a plan based on one or more smaller (25 GL or more) dams. None of them were economically sound, except when the magic of regional input-output multipliers was invoked.
In the light of Scott Morrison’s latest exercise in jettisoning unpopular commitments, in this case the proposal to raise the pension age to 70, I thought I would relink this piece on the Intergenerational Report, from the Abbott-Hockey era. The crucial observation is that, had the increase gone ahead, it would have cancelled out all of the increase in conditional life expectancy at pension age for women since the pension was introduced back in 1907, and most of the increase for men. The only real problem in retirement incomes policy is the lavish concessional treatment of superannuation.
The Liberals’ disastrous result in the recent Longman by-election obviously played a major role in bringing an effective end to Malcolm Turnbull’s Prime Ministership. But the lesson drawn from the outcome by nearly all political pundits, and particularly those on the political right seems to me to be totally unfounded.
The central claim is that the Liberals lost votes to One Nation, which more accurately reflected the views of their conservative basis. The corollary is that to win seats in Queensland the LNP needs to become more overtly racist, most obviously by elevating Peter Dutton to the leadership.
I won’t comment on the morality of this, but simply on the electoral mathematics. Let’s look at the electoral results for Longman, conveniently collected by Wikipedia. First, compare the by-election to the 2016 result. Obviously, the LNP vote collapsed. But what’s more striking is that the combined LNP-ONP coalition vote also fell by around 3 percentage points, while the Labor-Green coalition gained 5 per cent. The combined vote for each side was about 44 per cent. So, even if ONP preferences had flowed more strongly to the LNP, the outcome would have been very close.
The other part of the argument seems to be that Longman is representative of Australia, or at least Queensland as a whole. In reality, it’s classic One Nation territory*. In its first outing in the 1998 Queensland election, One Nation won the state seat of Caboolture (central to Longman), one of only a handful of wins in the South-east. In the Federal election the same year, One Nation got 18 per cent of the vote, more than this time around. That compares to a Queensland average of 14 per cent and a national average of 8 per cent. Interestingly, the One Nation vote in Dickson (now held by Dutton) was just 8 per cent.
The real problem is not that LNP voters as a group have suddenly become racists (or, at best, anti-anti-racists), but that the party’s members, activists and intellectual base have done so, but have had to conceal or blur the fact until relatively recently**. That’s why they are eager to adopt an interpretation of the Longman outcome that justifies them in coming out.
* To be absolutely clear, I don’t mean that most, or even a large minority of residents of Longman are racists or Hanson fans. Rather, whereas the average proportion of such people in Australia is around 10 per cent, in Longman its closer to 20. Whenever a Hanson-type candidate looks plausible, they can expect to get a fair few of those votes.
** I tried to think of someone who could reasonably be described as a small-l liberal in the way this term was once used. My best candidates were Peter van Onselen and Chris Berg, neither of whom really fit the bill in the way that, say, Ian McPhee did. Any others?