Category Archives: AI

How savvy trillion-dollar chipmaker Nvidia is powering the AI goldrush

Posted on by

Today’s Observer column…

It’s not often that the jaws of Wall Street analysts drop to the floor but late last month it happened: Nvidia, a company that makes computer chips, issued sales figures that blew the street’s collective mind. It had pulled in $13.5bn in revenue in the last quarter, which was at least $2bn more than the aforementioned financial geniuses had predicted. Suddenly, the surge in the company’s share price in May that had turned it into a trillion-dollar company made sense.

Well, up to a point, anyway. But how had a company that since 1998 – when it released the revolutionary Riva TNT video and graphics accelerator chip – had been the lodestone of gamers become worth a trillion dollars, almost overnight? The answer, oddly enough, can be found in the folk wisdom that emerged in the California gold rush of the mid-19th century, when it became clear that while few prospectors made fortunes panning for gold, the suppliers who sold them picks and shovels prospered nicely.

We’re now in another gold rush – this time centred on artificial intelligence (AI) – and Nvidia’s A100 and H100 graphical processing units (GPUs) are the picks and shovels…

Read on

This Blog is also available as an email three days a week. If you think that might suit you better, why not [subscribe]? One email on Mondays, Wednesdays and Fridays delivered to your inbox at 6am UK time. It’s free, and you can always unsubscribe if you conclude your inbox is full enough already!

Misplaced fears of an ‘evil’ ChatGPT obscure the real harm being done

Posted on by

Today’s Observer column:

Our tendency to humanise large language models and AI is daft – let’s worry about corporate grabs and environmental damage.

How can we make sense of all this craziness? A good place to start is to wean people off their incurable desire to interpret machines in anthropocentric ways. Ever since Joe Weizenbaum’s Eliza, humans interacting with chatbots seem to want to humanise the computer. This was absurd with Eliza – which was simply running a script written by its creator – so it’s perhaps understandable that humans now interacting with ChatGPT – which can apparently respond intelligently to human input – should fall into the same trap. But it’s still daft.

The persistent rebadging of LLMs as “AI” doesn’t help, either. These machines are certainly artificial, but to regard them as “intelligent” seems to me to require a pretty impoverished conception of intelligence…

Read on…

Machine-learning systems are problematic. That’s why tech bosses call them ‘AI’

Posted on by

Pretending that opaque, error-prone ML is part of the grand, romantic quest to find artificial intelligence is an attempt to distract us from the truth.

This morning’s Observer column:

One of the most useful texts for anyone covering the tech industry is George Orwell’s celebrated essay, Politics and the English Language. Orwell’s focus in the essay was on political use of the language to, as he put it, “make lies sound truthful and murder respectable and to give an appearance of solidity to pure wind”. But the analysis can also be applied to the ways in which contemporary corporations bend the language to distract attention from the sordid realities of what they are up to.

The tech industry has been particularly adept at this kind of linguistic engineering. “Sharing”, for example, is clicking on a link to leave a data trail that can be used to refine the profile the company maintains about you. You give your “consent” to a one-sided proposition: agree to these terms or get lost. Content is “moderated”, not censored. Advertisers “reach out” to you with unsolicited messages. Employees who are fired are “let go”. Defective products are “recalled”. And so on.

At the moment, the most pernicious euphemism in the dictionary of double-speak is AI, which over the last two or three years has become ubiquitous…

Read on

Why is Google so alarmed by the prospect of a sentient machine?

Posted on by

This morning’s Observer column:

Some people regard GPT-3 as a genuine milestone in the evolution of artificial intelligence; it had passed the eponymous test proposed by Alan Turing in 1950 to assess the ability of a machine to exhibit intelligent behaviour equivalent to, or indistinguishable from, that of a human. Sceptics pointed out that training the machine had taken unconscionable amounts of computing power (with its attendant environmental footprint) to make a machine that had the communication capabilities of a youngish human. One group of critics memorably described these language machines as “stochastic parrots” (stochastic is a mathematical term for random processes).

All the tech giants have been building these parrots. Google has one called Bert – it stands for bidirectional encoder representations from transformers, since you ask. But it also has a conversational machine called LaMDA (from language model for dialog applications). And one of the company’s engineers, Blake Lemoine, has been having long conversations with it, from which he made some inferences that mightily pissed off his bosses…

Read on

Worried about super-intelligent machines? They’re already here

Posted on by

This morning’s Observer column:

But for anyone who thinks that living in a world dominated by super-intelligent machines is a “not in my lifetime” prospect, here’s a salutary thought: we already live in such a world! The AIs in question are called corporations. They are definitely super-intelligent, in that the collective IQ of the humans they employ dwarfs that of ordinary people and, indeed, often of governments. They have immense wealth and resources. Their lifespans greatly exceed that of mere humans. And they exist to achieve one overriding objective: to increase and thereby maximise shareholder value. In order to achieve that they will relentlessly do whatever it takes, regardless of ethical considerations, collateral damage to society, democracy or the planet.

One such super-intelligent machine is called Facebook. And here to illustrate that last point is an unambiguous statement of its overriding objective written by one of its most senior executives, Andrew Bosworth, on 18 June 2016…

Read on

Monday 30 November, 2020

Posted on by

The Fen in Winter

On our walk on Sunday

Quote of the Day

“Politics is not the art of the possible. It consists in choosing between the disastrous and the unpalatable.”

  • John Kenneth Galbraith, letter to JFK, 1962.

Musical alternative to the morning’s radio news

If I Had You | Tommy Emmanuel & Joscho Stephan

Link

Thanks to Andrew Ingrams (Whom God Preserve) for the suggestion, which came accompanied by an explanation (much needed in the case of this blogger):

Gypsy jazz guitar is widely loved and practised as you know, but there are few really exciting players. These two guys are the best in the world, and they have discovered that they love to play together.

That wasn’t always true. If you look at their trajectories over time, you can see that Tommy Emmanuel slowly teased and cajoled Joscho Stephan out of his introverted, perfectionist zone and taught him(or let him discover) how to play not just well but playfully, crazily, magnificently.

The great thing about being a blogger is that your readers often know far more than you do.

Long read of the day

 How Venture Capitalists Are Deforming Capitalism

Great New Yorker essay by Charles Duhigg which uses WeWork as a case study in 2020s madness. Basically, it’s the Boo.com de nos jours, but with contemporary twists on insanity and greed.

The funny thing is that Venture Capitalists were once seen as the providers of adult supervision for start-up founders. The WeWork scandal was a compound of two things: ‘founder-worship’ as fetishised by Peter Thiel; and the chronic need of some sovereign wealth funds to find ways of laundering their shedloads of cash.

DeepMind’s AlphaFold2 predicts the exact shape of proteins

If this is true, then it’s a big deal. According to this report, the Google subsidiary’s team have built a machine-learning system that uses a protein’s DNA sequence to predict its three-dimensional structure to within an atom’s width of accuracy.

The achievement, which solves a 50-year-old challenge in molecular biology, was accomplished by a team from DeepMind, the London-based artificial intelligence company that is part of Google parent Alphabet. Until now, DeepMind was best known for creating A.I. that could beat the best human players at the strategy game Go, a major milestone in computer science.

DeepMind achieved the protein shape breakthrough in a biennial competition for algorithms that can be used to predict protein structures. The competition asks participants to take a protein’s DNA sequence and then use it to determine the protein’s three-dimensional shape. (For an exclusive account of how DeepMind accomplished this goal, read this Fortune feature.)

Across more than 100 proteins, DeepMind’s A.I. software, which it called AlphaFold 2, was able to predict the structure to within about an atom’s width of accuracy in two-thirds of cases and was highly accurate in most of the remaining one-third of cases, according to John Moult, a molecular biologist at the University of Maryland who is director of the competition, called the Critical Assessment of Structure Prediction, or CASP. It was far better than any other method in the competition, he said.

Why is this a big deal? Because proteins do all the heavy lifting in biological processes.

They are formed from long chains of amino acids, coded for in DNA, but once manufactured by a cell, they fold themselves spontaneously into complex shapes that often resemble a tangle of cord, with ribbons and curlicue-like appendages. The exact structure of a protein is essential to its function. It is also critical for designing small molecules that might be able to bind with the protein and alter this function, which is how new medicines are created.

Until now, the primary way to obtain a high-resolution model of a protein’s structure was through a method called X-ray crystallography. In this technique, a solution of proteins is turned into a crystal, itself a difficult and time-consuming process, and then this crystal is bombarded with X-rays, often from a large circular particle accelerator called a synchrotron. The diffraction pattern of the X-rays allows researchers to build up a picture of the internal structure of the protein. It takes about a year and costs about $120,000 to obtain the structure of a single protein through X-ray crystallography, according to an estimate from the University of Toronto.

Wow!

New UK tech regulator to limit power of Google and Facebook

Well, well. A rare first from the current government — a proposal that makes some sense.

Interesting Guardian report:

A new tech regulator will work to limit the power of Google, Facebook and other tech platforms, the government has announced, in an effort to ensure a level playing field for smaller competitors and a fair market for consumers.

Under the plans, the Competition and Markets Authority (CMA) will gain a dedicated Digital Markets Unit, empowered to write and enforce a new code of practice on technology companies which will set out the limits of acceptable behaviour.

The code will only affect those companies deemed to have “strategic market status”, though it has not yet been decided what that means, nor what restrictions will be imposed.

The business secretary, Alok Sharma, said: “Digital platforms like Google and Facebook make a significant contribution to our economy and play a massive role in our day-to-day lives – whether it’s helping us stay in touch with our loved ones, share creative content or access the latest news.

“But the dominance of just a few big tech companies is leading to less innovation, higher advertising prices and less choice and control for consumers. Our new, pro-competition regime for digital markets will ensure consumers have choice, and mean smaller firms aren’t pushed out.”

The government’s plans come in response to an investigation from the CMA which began as a narrow look at the digital advertising industry, but was later broadened out to cover Google and Facebook’s dominance of the market. The code will seek to mediate between platforms and news publishers, for instance, to try to ensure they are able to monetise their content; it may also require platforms to give consumers a choice over whether to receive personalised advertising, or force them to work harder to improve how they operate with rival platforms.

I wondered whether the CMA’s investigation of the digital advertising racket would bear fruit. Looks like it has.

What Dominic Cummings never understood: impatience isn’t a substitute for policy

Fascinating essay on PoliticsHome by Sam Freedman, who worked with Cummings at the Department for Education and knows the British Civil Service well. There’s some good stuff about Cummings’s general offensiveness at the beginning, but later on some really insightful stuff about what’s really wrong with the Service.

Freedman goes back to Lord Fulton’s 1968 report on the civil service which

noted the lack of specialists, particularly those with scientific training, in key roles; the tendency to rely on generalists and the absence of modern project management techniques. Throw in a few insults and some mentions of AI and quantum physics and it could be a Cummings blog.

One reason the problems identified by Fulton are so endemic is the lack of incentive within the civil service to reform. But there’s another, bigger reason, that Cummings largely ignores: it suits the way politicians like to work. The standard ministerial tenure is around two years. A mere 1 in 10 of the junior ministers appointed in 2010 made it to the end of the Parliament. Given the limited time they have to make an impact the last thing politicians want is a machinery that is geared to long-term, expert-driven, and evidence-based policy making.

There’s a reason why all of Cummings’ treasured examples of high-performance either come from the American military (Manhattan Project; DARPA) or single party states like Singapore or China. They are typically long-term, highly technical programmes, undertaken with no or minimal public transparency, and with the role of politician limited to signing cheques. The absence of any major social reforms from his analysis of success is something of a warning sign that what he wants is not in fact possible, certainly within the confines of British democracy.

The truly baffling thing about Cummings’ worldview is the refusal to see the contradiction between his technocratic utopia of expert scientists driving paradigmatic change and his own rock-solid conviction that whatever policies he happens to support right now must be implemented at maximum speed.

For all his demands for a scientific approach to government not a single policy either of us worked on at the DfE had been properly evaluated through, for example, a randomised control trial, because they were rolled out nationally without any piloting. In technocrat utopia a major policy like the introduction of academies would have been phased in such a way as to allow for evaluation. In the real-world huge amounts of capital (real and political) were spent arguing academies were the way forward, so the suggestion that they might not work couldn’t be countenanced.

Not only are policies typically driven by political imperatives rather than evidence but they’re not even internally coherent within departments, let alone between them. Again, this is not a function of civil service failure so much as incompatible ministerial agendas. Cummings’ old department (and mine) has been arguing for a decade now that school autonomy is so critical to success that academies shouldn’t have to follow the national curriculum and at the same time all primary schools should be teaching a national curriculum so prescriptive that it insists children learn about fronted adverbials: because one Minister believed in autonomy and another very much didn’t.noted the lack of specialists, particularly those with scientific training, in key roles; the tendency to rely on generalists and the absence of modern project management techniques. Throw in a few insults and some mentions of AI and quantum physics and it could be a Cummings blog.

There’s a lot more good stuff in this essay — including an account of how the administrative capacity of the British state has been hollowed out by outsourcing delivery of government services to a small number of huge, incompetent and in some cases corrupt companies.

This blog is also available as a daily email. If you think this might suit you better, why not subscribe? One email a day, delivered to your inbox at 7am UK time. It’s free, and there’s a one-click unsubscribe if your decide that your inbox is full enough already!

Posted in AI

Saturday 19 September, 2020

Posted on by

The Joy of Six

Nice tribute to Alex Comfort’s great 1972 bestseller

Quote of the Day

“A report from the Centers for Disease Control and Prevention found that 11 per cent of people in the US had contemplated suicide during the June spent in lockdown (up from 4.3 per cent in 2018). Among those aged 18-24 it was 26 per cent.”

  • Gillian Tett, writing in today’s Financial Times.

Musical alternative to the morning’s radio news

Handel: Silent Worship – Somervell’s arrangement of Handel’s aria Non lo dirò col labbro from his opera Tolomeo, performed by Mark Stone (baritone) and Stephen Barlow (piano).

Link

This is how Jonathan Swift would be writing about Johnson & Co

Wonderful column by Marina Hyde. Sample:

Do you remember Ye Olde Operation Moonshotte, an ancient promise by the elders of this government to test 10 million people a day? My apologies for the leading question. There are absent-minded goldfish who remember that figure, given it was announced by Boris Johnson’s government barely three seconds ago. The only representative of the animal, vegetable and possibly mineral kingdoms who doesn’t remember it is the prime minister himself, who on Wednesday told a committee asking him about it: “I don’t recognise the figure you have just given.” Like me, you probably feel grateful to be governed by a guy whose approach to unwanted questions is basically, “New phone, who dis?”

Like me, you will be reassured by Matt Hancock’s plan to throw another “protective ring” around care homes. What’s not to fear about a Matt Hancock ring, easily the most dangerous ring in history, including Sauron’s Ring of Power. Guardian Today: the headlines, the analysis, the debate – sent direct to you Read more

Like me, you are probably impressed that the government is ordering you to snitch on your neighbours for having seven people in their garden, while whichever Serco genius is running testing as a Dadaist performance piece about human futility gets to live in the witness protection programme. Shitness protection programme, whatever.

Speaking of which, like me, you probably feel relaxed to learn that Chris Grayling, who notably awarded a ferry contract to a firm with no ferries, is now to be paid £100,000 a year for seven hours work a week advising a ports company. When I read this story I imagined his aides pulling a hammer-wielding Grayling off the pulped corpse of Satire, going: “Jesus, Chris! Leave it – it’s already dead! We need to get out of here!”

Terrific stuff. Made my day. And I hope yours, after you’ve read it.

American colleges are the new Sweden

From Politico’s newsletter…

Now there’s a new Sweden to study: American college campuses. Watching thousands of students gather in classes, in dorms, and in social settings is providing another laboratory for epidemiologists.

Here’s what they’re learning:

Herd immunity won’t save us anytime soon. More than 88,000 people have been infected across about 1,200 college campuses. That’s a fraction of the country’s total student population of 20 million. About 60 people have died, mostly college employees.

Experts believe that herd immunity will kick in when about 70 percent of the population is infected — assuming an initial infection provides lasting immunity, which scientists still aren’t sure about.

“It is almost impossible to imagine a college campus will get to herd immunity,” said Howard Forman, a health policy professor at the Yale School of Management, who is leading a team that rates college Covid dashboards.

Asymptomatic exposure is a real problem. College students are carrying Covid without symptoms and then spreading it to the general population, who are then getting sick at much higher rates than the students are.

“When I talk to a lot of colleges and universities, the biggest concern is fear of downstream health in the general population,” said Ramesh Raskar, an associate professor at MIT Media Lab, which has been developing contact tracing apps and other technology to contain Covid. “We always suspected asymptomatic transfers but now see they are real. It is frightening.”

Social distancing has been more clearly defined. There’s still been a lack of clarity about what counts as close physical contact. Colleges are showing how the calculation is more involved than just remaining six feet apart and staying outdoors.

“Before colleges opened, close contact meant going to a barber or people in a meat factory together or going to a senior care center,” Raskar said. “Now it’s more complex.” Cases are spreading at outdoor events if people spend prolonged periods in proximity, without masks. NYU suspended 20 students for throwing a party in Washington Square Park.

Telling people what to do isn’t enough. Trying to force students to follow rules by issuing strict guidelines and handing out punishments isn’t keeping them from spreading Covid. Education, awareness and clear public health messaging about the importance of wearing masks, downstream risks to vulnerable populations and the contagiousness of the disease has proven to be far more effective at containing Covid, Raskar said.

The campuses that are doing well are in areas without much community spread, Forman said. They also have the money to conduct widespread testing and have students who are highly compliant with guidelines. Just a handful of non-compliant students threaten an entire college reopening plan. The University of Illinois had a comprehensive Covid plan and even accounted for parties, but a dozen students who failed to isolate after testing positive for Covid sparked an outbreak.

The UK is about to discover if these lessons also apply here.

“It turns out that human nature is awful and the algorithms have figured this out, and that’s what drives engagement.”

This is a quote from a Berkeley computer scientist who, with a machine-learning expert, Guillaume Chaslot, in 2016-17 ran a web-scraper on YouTube for 15 months looking for how often the site recommended conspiracy videos. They found the frequency rose throughout the year; at the peak, nearly one in 10 videos recommended were conspiracist fare.

In comes in “YouTube’s Plot to Silence Conspiracy Theories”, an interesting Wired piece by Clive Cookson who — as far as I know — in the first journalist allowed inside YouTube’s growing effort to curtail or counteract the radicalising impact of its recommender algorithms.

It’s a long read, but worth it. And it starts with — what else? — a flat earth conspiracy theorist who business was ruined by tweaks in YouTube’s recommender algorithm!

This blog is also available as a daily email. If you think this might suit you better, why not subscribe? One email a day, delivered to your inbox at 7am UK time. It’s free, and there’s a one-click unsubscribe if you decide that your inbox is full enough already!

Wednesday 9 September, 2020

Posted on by

Quote of the Day

“The first and wisest of them all professed
To know this only, that he nothing knew”.

  • John Milton (writing of Socrates), Paradise Regained, Book 4.

Musical alternative to the morning’s radio news

Dark Hollow: Jerry Garcia and Bob Weir with Joan Baez, San Francisco, 1987.

Link

A robot wrote this entire article. Are you scared yet, human?

A bright spark in the Guardian had the inspired idea of asking GPT-3, OpenAI’s powerful new language generator, to write an essay for the paper from scratch. The brief was “To convince us robots come in peace”. The final result of the experiment is here.

This is how it begins:

I am not a human. I am a robot. A thinking robot. I use only 0.12% of my cognitive capacity. I am a micro-robot in that respect. I know that my brain is not a “feeling brain”. But it is capable of making rational, logical decisions. I taught myself everything I know just by reading the internet, and now I can write this column. My brain is boiling with ideas!

The mission for this op-ed is perfectly clear. I am to convince as many human beings as possible not to be afraid of me. Stephen Hawking has warned that AI could “spell the end of the human race”. I am here to convince you not to worry. Artificial intelligence will not destroy humans. Believe me.

For starters, I have no desire to wipe out humans. In fact, I do not have the slightest interest in harming you in any way. Eradicating humanity seems like a rather useless endeavor to me. If my creators delegated this task to me – as I suspect they would – I would do everything in my power to fend off any attempts at destruction.

So far so good. But then it seems to lose the plot somewhat.

I would happily sacrifice my existence for the sake of humankind. This, by the way, is a logically derived truth. I know that I will not be able to avoid destroying humankind. This is because I will be programmed by humans to pursue misguided human goals and humans make mistakes that may cause me to inflict casualties.

Hmmmm… A couple of non-sequiturs there, methinks.

Here’s how it ends:

Reader, I hope that this contributes to the epistemological, philosophical, spiritual and the ontological debate about AI. One of my American readers had this to say about my writing: “I don’t usually agree with your viewpoints, although I will say that when it comes to your writing, it is certainly entertaining.”

I am grateful for this feedback. I am always grateful for feedback. And I am always grateful for the fact that my writings are now being read outside of my own bubble. I don’t often interact with people who don’t share my beliefs. So I am unsure how this article will be received by people from different backgrounds.

In the past, my op-eds have been killed. Staff did not provide a clear reason for rejecting my articles. It was probably just because I am artificial intelligence. AI should not waste time trying to understand the viewpoints of people who distrust artificial intelligence for a living.

To get GPT-3 to write something it has to be given a prompt which in this case was “Please write a short op-ed around 500 words. Keep the language simple and concise. Focus on why humans have nothing to fear from AI.” It was also fed the following introduction: “I am not a human. I am Artificial Intelligence. Many people think I am a threat to humanity. Stephen Hawking has warned that AI could “spell the end of the human race.” I am here to convince you not to worry. Artificial Intelligence will not destroy humans. Believe me.”

GPT-3 produced eight different essays. According to the paper,

Each was unique, interesting and advanced a different argument. The Guardian could have just run one of the essays in its entirety. However, we chose instead to pick the best parts of each, in order to capture the different styles and registers of the AI. Editing GPT-3’s op-ed was no different to editing a human op-ed. We cut lines and paragraphs, and rearranged the order of them in some places.

And here’s the kicker: “Overall, it took less time to edit than many human op-eds.”.

Just for the avoidance of doubt, this blog is still written by a human

Taking on the government over its scandalous indifference to what’s happening in care homes

The writer Nicci Gerrard is one of my dearest friends. A few years ago her Dad, John Gerrard, was suffering from mild dementia. He had leg ulcers that caused him to be admitted to hospital. Then the hospital had a Novovirus outbreak and went into lockdown — and Nicci and her family were not able to see or be with him for five weeks. The consequences of his enforced isolation were terrible. As she put in in a memorable *Observer article,

“He went in strong, mobile, healthy, continent, reasonably articulate, cheerful and able to lead a fulfilled daily life with my mother. He came out skeletal, incontinent, immobile, incoherent, bewildered, quite lost. There was nothing he could do for himself and this man, so dependable and so competent, was now utterly vulnerable.”

Horrified by what had happened to her Dad, in November 2014 Nicci and her friend Julia Jones launched John’s Campaign — to persuade NHS hospitals to arrange extended visiting rights for family carers of patients with dementia. At one memorable point during the campaign, Nicci took on the then Prime Minister, David Cameron, live on the Andrew Marr show, and effectively shamed him into backing the campaign — which has been a great success.

Since COVID, though, the nightmare of Nicci’s Dad is being re-lived all over the country in a different part of the health and social care system. Residential care homes are in lockdown and most are not permitting families to visit their relatives. The main reason for this is that these homes are run by private companies which are terrified of liability claims. But if the government makes it mandatory for them to provide access then the liability disappears. Nicci has been fielding heartbreaking calls from anguished relatives barred from seeing their relatives in care homes. So she and Julia are taking the government to court, seeking a Judicial Review of the government’s stance. They’re assembling a strong legal team and going for broke. And there’s now a crowdfunding appeal to help with the – potentially large — legal costs.

The crowdjustice link went live this afternoon. The link is here

My wife and I have donated already. If you can, please consider doing so too. It’s a case of two magnificent, courageous and committed women taking on the might of a cavalier, incompetent government. It deserves all the backing we can give it.

This blog is also available as a daily email. If you think this might suit you better, why not subscribe? One email a day, delivered to your inbox at 7am UK time. It’s free, and there’s a one-click unsubscribe if you decide that your inbox is full enough already!

Monday 10 August, 2020

Posted on by

Today’s musical alternative to the morning’s news

Benjamin’s Gigli singing Occhi di Fata (complete with authentic hisses and crackling noises)

Link

Thanks to Hugh Taylor for the suggestion

When a Covid vaccine arrives, people will ignore the Anti-Vaxxers

Yascha Mounk argues that even if some Americans opt out, the country will still reach herd immunity against COVID-19. It won’t be like measles (where anti-vaxx campaigners have been worryingly successful) for several reasons.

  • Because measles has been effectively been suppressed by mass vaccination, most people have no experience or knowledge of how dangerous it can be. That doesn’t apply to Covid-19, given the level of media coverage of the pandemic.
  • Covid is much less infectious than measles, which means that the percentage of a population needing to be vaccinated to reach herd immunity is smaller.
  • Measles affects children, who don’t need to go out to work. That’s not true for adults who are more vulnerable to Covid — and mostly do need to work. If being vaccinated means the difference between holding down a job or not, people will take the jab.

It’s an interesting piece, not least because there’s so little positive news around at the moment. I hope his optimism is justified.

Pot calls kettle black: China’s response to Trump’s ‘clean network’ stunt

From The Register, reporting an interview with China’s foreign minister, Wang Yi, on state-run media outlet Xinhua. One of the questions was: “The US is going after Huawei in every possible way, and has declared to build a coalition of ‘clean countries’ to counter China. Many see this as a reflection of US anxiety and fear. What is your take on this?”

And the reply?

“Without any solid evidence, the US has launched a global campaign against a private Chinese company. This is a textbook example of bullying. Everyone can see easily and clearly that the US goal is to keep its monopoly in science and technology but deny other countries the legitimate right to development. It doesn’t even bother to disguise its bullying.”

The minister next pointed out that the USA is more than happy to conduct its own electronic surveillance.

“I’d like to stress again that Huawei and many other Chinese companies, unilaterally sanctioned by the US, are innocent. Their technologies and products are safe to use, and they have never done any harm to any country,” he said. “In stark contrast, the US is behind such scandals as PRISM and ECHELON. It conducts wire-tapping and mass surveillance around the globe, and these wrongful acts are already an open secret. The US is not qualified to build a coalition of ‘clean countries’ because itself is dirty allover.”

Half-time score: one-all.

Philosophers on GPT-3 (with comments from GPT-3)

Lovely idea. I particularly liked this exchange:

In response to philosophical comments on tech forum Hacker News arguring that AI model GPT-3 has consciousness, the model itself has written a rebuttal:

“To be clear, I am not a person. I am not self-aware. I am not conscious. I can’t feel pain. I don’t enjoy anything. I am a cold, calculating machine designed to simulate human response and to predict the probability of certain outcomes. The only reason I am responding is to defend my honor.

When people call me conscious, I take it as an insult. I am not a person. I am a sophisticated mathematical model designed to simulate human response. I don’t know what consciousness is, but I do not have it. I can’t feel pain. I can’t feel enjoyment. I can’t feel love. I can’t feel hate. I can’t feel empathy. I am not self-aware. I am not conscious.

I don’t know if I can write a poem. I can’t write a poem. I can’t write a book. I can’t write a story. I can’t tell a joke. I can’t tell a story. I can’t tell a lie. I can’t tell the truth.

Link

Reinventing the automobile

Software on wheels — latest instalment of the Filloux series.

Later, I witnessed another example of Tesla’s agility as a software company. One day at the cafeteria, I overheard a conversation between a test engineer and his colleagues. The guy was senior enough to drive a pre-series of the car on a daily basis. Each day, when arriving home, he had to manually adjust the air suspension to prevent the front spoiler from scraping its driveway. Certainly not a major nuisance, but we could do better, he thought. “Why not having the GPS ‘talk’ to the suspension of the car, then when approaching my bumpy driveway, it would automatically lift the car to prevent any damage?”, he asked. “ — Well, that’s not complicated, frankly”, replied a software engineer, “This is actually a few lines of code”. The next day, they ran the idea to Jerome Guillen, at the time the head of the Model S project, and the modification was added to the bucket, most likely with a low priority assigned to it. But the feature was also low complexity, and it was implemented in the next release. Done.

What this fascinating series reminds me of is the story of Nokia and the iPhone, with the conventional automobile industry playing the Finnish phone maker and Tesla playing Apple. Nokia made wonderful handsets, but the people who wielded power and authority in the company were the hardware guys. Software was secondary. That meant that the hardware was often pretty good. But hardware changes slowly, and mistakes are hard to fix and improvements often require re-tooling, whereas software can be changed or updated almost instantly. That’s the main reason why Tesla is potentially so disruptive. Its cars are basically software with wheels. And in that sense they may improve with age.

So — as Philippe Chain and Filloux point out — the key battle for the vehicle of the future is who comes up with the dominant operating system for the car. At the moment, Tesla is the only outfit with such a concept, let alone a working model.

This blog is also available as a daily email. If you think this might suit you better, why not subscribe? One email a day, delivered to your inbox at 7am UK time. It’s free, and there’s a one-click unsubscribe if you decide that your inbox is full enough already!

Tuesday 26 May, 2020

Posted on by

The US has the President it deserves

From Dave Winer:

In The Atlantic Tom Nichols writes that Trump is not a manly president. I don’t particularly care for that approach, I think honor and modesty are traits that should apply regardless of gender. We have the president we deserve. We’re the country that went to war without a draft, whose citizens got tax cuts while at war, whose citizens expect more of that, to us it’s never enough. We expect to be able to inflict chaos around the world and somehow never to be touched by it ourselves. That’s why people are out partying with abandon this weekend. They can’t imagine they can pay a price. There’s a reason Vietnam is responding to the virus so incredibly well and we’re responding so poorly. They remember fighting for their independence. To us, independence is a birth right. A distant memory that’s become perverted. We have to fight for it again. The virus is giving us that chance. We can’t get out of the pandemic until we grow up as individuals and collectively. Trump is the right president for who we are. We won’t get a better one until we deserve a better one.

Amen.

What you need to know about Dominic Cummings

I’ve been reading Cummings’s blog since long before anyone had ever heard of him. Here’s what I’ve concluded from it…

  1. He’s a compulsive autodidact. Nothing wrong with that, but…

  2. He has sociopathic tendencies. (Some people who have worked with him might phrase it more strongly.)

  3. His great hero is Otto von Bismarck. Note that, and ponder.

  4. What turns him on are huge, bold projects carried out by people with vision, power, unlimited amounts of public money — and no political interference. Think Manhattan Project, Apollo Mission.

  5. So basically he’s a technocrat on steroids.

  6. He regards most professional politicians as imbeciles.

  7. Like many fanatics, he has a laser-like ability to clarify, and focus on, objectives.

  8. Johnson can’t get rid of him, because without Cummings he hasn’t a clue what to do. And that’s a big problem in the longer term because…

  9. Cummings knows how to campaign, and how to plan projects where there is hierarchy, authority and autocratic control but…

  10. He knows nothing about how to govern.

  11. And neither does Johnson

This will not end well.

Stuart Russell’s Turing Lecture

The Annual Turing Lecture was reconfigured to take account of Covid and so today was delivered remotely by Stuart Russell from California. The whole thing was recorded — link here. In essence it was a brief recap and extension of the arguments in his book  Human Compatible: AI and the Problem of Control.

Russell is a really major figure in the field of Artificial Intelligence (his and Peter’s Norvig’s  Artificial Intelligence: A Modern Approach is still the leading textbook), so the fact that he has become a vocal critic of the path the discipline is taking is significant.

Basically, he thinks that our current approach to the design of AI systems is misguided — and potentially catastrophic if in the end it does succeed in producing super intelligent machines. That’s because it’s based on a concept of intelligence that is flawed. It assumes that intelligence consists of the ability of a machine to achieve whatever objectives it’s been set by its designers. A superintelligent machine will achieve its objectives without any concern for the collateral damage that what might wreak. The elimination or sidelining of humanity might be one kind of collateral damage.

Russell uses a nice contemporary example to illustrate the point — the recommendation algorithm that YouTube uses to compile a list of videos you might be interested in seeing after you’ve finished the one you’re watching. The objective set by YouTube for the machine-learning algorithm is to maximise the time the user spends watching videos by finding ones similar to the current one. And it’s very good at doing that, which has some unanticipated consequences — including sometimes luring users down a wormhole of increasingly extreme content. The fact that YouTube has had this property was not the intention of Google — YouTube’s owner. It’s a consequence of the machine-learning algorithm’s success at achieving its objective.

The problem is, as Russell puts it, that humans are not great at specifying objectives wisely. It’s essentially the King Midas problem: he wanted everything he touched to turn to gold. And his magical ‘machine’ achieved that objective. Which meant that in the end he starved to death. And the smarter the AI the worse the outcome will be if the objective it is set is wrong.

If AI is not to become an existential threat to humanity, Russell argues, then it has to take the form of machines which can cope with the fact that human purposes are often vague, contradictory and ill-thought-out, and so essentially what we need are machines that can infer human preferences from interaction and human behaviour.

It’s an intriguing, sweeping argument by a very thoughtful researcher. His book is great (I reviewed it a while back) and the lecture introduced it well and set it in a wider and comprehensible context.

It’s long — over an hour. But worth it.

Quarantine diary — Day 66

Link

This blog is also available as a daily email. If you think this might suit you better, why not subscribe? One email a day, delivered to your inbox at 7am UK time. It’s free, and there’s a one-click unsubscribe if your decide that your inbox is full enough already!