Start Up No.981: hot v cool v Trump v AOC, AI v Alzheimer’s, the oven that knows what it’s cooking, Firefox kills Flash, and more

Recognise this place? Its average internet speed is faster than France, Canada or the UK. CC-licensed photo by mariusz kluzniak on Flickr.

You can sign up to receive each day’s Start Up post by email. You’ll need to click a confirmation link, so no spam.

A selection of 9 links for you. Still here? I’m @charlesarthur on Twitter. Observations and links welcome.

Hot Trump. Cool @aoc • Medium

Jeff Jarvis:


I’ve been rereading a lot of Marshall McLuhan lately and I’m as confounded as ever by his conception of “hot” vs. “cool” media. And so I decided to try to test my thinking by comparing the phenomena of Donald Trump and Alexandria Ocasio-Cortez at this millennial media wendepunkt, as text and television give way to the net and whatever it becomes. I’ll also try to address the question: Why is @aoc driving the GOP mad?

…As TV became hotter [in McLuhan terms] — as it became high-definition — it found its man in Trump, who is as hot and unsubtle as a thermonuclear blast. Trump burns himself out with every appearance before crowds and cameras, never able to go far enough past his last performance — and it is a performance — to find a destination. He is destruction personified and that’s why he won, because his voters and believers yearn to destroy the institutions they do not trust, which is every institution we have today. Trump then represents the destruction of television itself. He’s so hot, he blew it up, ruining it for any candidate to follow, who cannot possibly top him on it. Kennedy was the first cool television politician. Obama was the last cool TV politician. Trump is the hot politician, the one who then took the medium’s every weakness and nuked it. TV amused itself to death.

Alexandria Ocasio-Cortez was not a candidate of television or radio or text because media — that is, journalists — completely missed her presence and success, didn’t cover her, and had to trip over each other to discover her long after voters had. How did voters discover her? How did she succeed? Social media: Twitter, Facebook, Instagram, YouTube….

I think McLuhan’s analysis here would be straightforward: Social media are cool [media forms]. Twitter in particular is cool because it provides such low-fidelity and requires the world to fill in so much, not only in interpretation and empathy but also in distribution (sharing). And Ocasio-Cortez herself is cool in every definition.

…Yes, she shoots at her opponents, but like a sniper, always from her position, her platform.


I found this a fascinating analysis. Like Jeff, I’ve tried for years to comprehend McLuhan’s declensions; this one makes it understandable. And I love the “sniper” line.
link to this extract

Artificial intelligence can detect Alzheimer’s disease in brain scans six years before a diagnosis • UC San Francisco

Dana Smith:


glucose PET scans are much more common and cheaper, especially in smaller health care facilities and developing countries, because they’re also used for cancer staging.

Radiologists have used these scans to try to detect Alzheimer’s by looking for reduced glucose levels across the brain, especially in the frontal and parietal lobes of the brain. However, because the disease is a slow progressive disorder, the changes in glucose are very subtle and so difficult to spot with the naked eye.

To solve this problem, Sohn applied a machine learning algorithm to PET scans to help diagnose early-stage Alzheimer’s disease more reliably.

“This is an ideal application of deep learning because it is particularly strong at finding very subtle but diffuse processes. Human radiologists are really strong at identifying tiny focal finding like a brain tumor, but we struggle at detecting more slow, global changes,” says Sohn. “Given the strength of deep learning in this type of application, especially compared to humans, it seemed like a natural application.”

To train the algorithm, Sohn fed it images from the Alzheimer’s Disease Neuroimaging Initiative (ADNI), a massive public dataset of PET scans from patients who were eventually diagnosed with either Alzheimer’s disease, mild cognitive impairment or no disorder. Eventually, the algorithm began to learn on its own which features are important for predicting the diagnosis of Alzheimer’s disease and which are not.

…The algorithm performed with flying colors. It correctly identified 92% of patients who developed Alzheimer’s disease in the first test set and 98% in the second test set. What’s more, it made these correct predictions on average 75.8 months – a little more than six years – before the patient received their final diagnosis.


Slightly scary. What do you do with a diagnosis like that?
link to this extract

Whirlpool WLabs oven can detect your food and cook it properly • CNBC

Todd Haselton:


A new countertop oven from Whirlpool’s WLabs can automatically detect the food you put in it, and cook it for the right time. It takes all of the guesswork out of cooking.

The oven was on display at CES last week where CNBC had a chance to see how it works. I tried to insert a bunch of fake asparagus in one demo and, in another, a tray of salmon. Sensors inside the oven were able to determine what I was trying to cook, and then proposed the right amount of cook time and temperature.

This is different than Amazon’s microwave, which knows how long to microwave certain foods, but can’t automatically detect what you’ve placed inside.

The oven can identify 50 different types of food using infrared sensors in the oven. For meat, a user inserts a probe into the filet so that the oven can cook it to your liking.


Only $800! Why do it?
link to this extract

Stop the presses: How a new publishing platform can help local news • Google blog

Jim Albrecht, product management director of search:


Shouldn’t doing great editorial work be enough?

We think so, and that’s why the Google News Initiative has partnered with Automattic/WordPress and invested $1.2m in its effort to create Newspack: a fast, secure, low-cost publishing system tailor-made to the needs of small newsrooms. Other funders include the Lenfest Institute for Journalism, The John S. and James L. Knight Foundation and Civil Media collectively contributing another $1m.

Journalists should be writing stories and covering their communities, not worrying about  designing websites, configuring CMSs, or building commerce systems. Their publishing platform should solve these problems for them. So while Newspack publishers will have access to all the plugins created by the WordPress developer community, the core product is not trying to be all things to all publishers. It is trying to help small publishers succeed by building best practices into the product while removing distractions that may divert scarce resources. We like to call it “an opinionated CMS:” it knows the right thing to do, even when you don’t.


Good that it’s built on WordPress – but will it get the security updates that WordPress gets? Will Google try to inveigle itself in, by doing updates?
link to this extract

Mozilla kills default support for Adobe Flash in Firefox 69 • Threatpost

Lindsay O’Donnell:


Firefox 69 will force users to manually install Adobe Flash as the plugin inches toward end of life.

Mozilla is disabling default support for Adobe’s Flash Player plugin in the latest upcoming version of its FireFox browser, marking the latest step in end-of-life for the infamous plugin.

The disabled default support means that Firefox users will now be required to manually enable Adobe Flash in Mozilla’s latest browser version, Firefox 69. More importantly, the change signals another step toward the end of Flash in general, as Mozilla and other popular browsers push the plugin off the radar.


Only 3.9% of sites use Flash now, compared to 28.5% in 2011. (Guessing that a big part of that fall is the rise in sites configured for mobile – where Flash isn’t installed.
link to this extract

Fortnite skins are key to the future of global trade • Bloomberg

Shawn Donnan:


Discussions about globalization—and its costs and benefits—often focus on physical goods such as steel beams, cars, or soybeans. The reality is that the integration of economies is increasingly a digital one that happens in invisible daily bursts—like the sessions in which far-flung armies of Fortnite players face off against each other on an imaginary island. “The digital economy is everywhere, and much of it is international without our even knowing it,” says Anupam Chander, a law professor and expert on digital trade at Georgetown University.

If we don’t always fully appreciate the scale of what’s going on, it’s because much of digital trade is not being captured in official statistics, says Susan Lund of the McKinsey Global Institute, the consultant’s in-house think tank. In a report, Lund and her co-authors documented an explosion in global data flows that they argued generated $2.8 trillion in economic output in 2014 alone and was doing more to benefit the world economy than the stalling international trade in physical goods.

In some cases, the World Trade Organization pointed out in a report last year, the rapid propagation of digital technologies has contributed to a false picture of globalization in retreat as shipping containers filled with hardcovers and DVDs are replaced by e-book downloads and streaming music.


I seem to recall that much the same was said about Second Life, and then World of Warcraft, and then EVE Online, but the general point is true – digital globalisation is changing things.
link to this extract

Only nuclear energy can save the planet • WSJ

Joshua Goldstein and Staffan Qvist:


the consumption of fossil fuels is growing quickly as poorer countries climb out of poverty and increase their energy use. Improving energy efficiency can reduce some of the burden, but it’s not nearly enough to offset growing demand.

Any serious effort to decarbonize the world economy will require, then, a great deal more clean energy, on the order of 100 trillion kilowatt-hours per year, by our calculations—roughly equivalent to today’s entire annual fossil-fuel usage. A key variable is speed. To reach the target within three decades, the world would have to add about 3.3 trillion more kilowatt-hours of clean energy every year.

Solar and wind power alone can’t scale up fast enough to generate the vast amounts of electricity that will be needed by midcentury, especially as we convert car engines and the like from fossil fuels to carbon-free energy sources. Even Germany’s concerted recent effort to add renewables—the most ambitious national effort so far—was nowhere near fast enough. A global increase in renewables at a rate matching Germany’s peak success would add about 0.7 trillion kilowatt-hours of clean electricity every year. That’s just over a fifth of the necessary 3.3 trillion annual target.

…So why isn’t everyone who is concerned about climate change getting behind nuclear power? Why isn’t the nuclear power industry in the U.S. and the world expanding to meet the rising demand for clean electricity? The key reason is that most countries’ policies are shaped not by hard facts but by long-standing and widely shared phobias about radiation.


There’s one other point: they’re pretty slow to build. But it’s true. We need more of them.
link to this extract

A first look at Twitter’s new beta app and its bid to remain ‘valuable and relevant’ • TechCrunch

Sarah Perez and Ingrid Lunden:


During the first beta, participants will try out new conversation features which offer color-coded replies to differentiate between responses from the original poster of the tweet, those from people you follow, and those from people you don’t follow.

In a development build of the beta app, Haider showed us what this looked like, with the caveat that the color scheme being used has been intentionally made to be overly saturated – it will be dialed down when the features launch to testers.

When you click into a conversation thread, the beta app will also offer visual cues to help you better find the parts of the thread that are of interest to you.

One way it’s doing so is by highlighting the replies in a thread that were written by people you follow on Twitter. Another change is that the person who posted the original tweet will also have their own replies in the thread highlighted.

In the build Haider showed us, replies from people she followed were shown in green, those from non-followers were blue, and her own replies were blue.


May also get rid of likes (maybe) and add a “status” field such as availability, location, whether you’re online “as on IM”, says the story – perhaps oblivious to the fact that Twitter was started with the intent of being an SMS equivalent of the IM status field.
link to this extract

Madagascar’s fast internet fuels outsourcing boom • Quartz Africa


There are now 233 BPO [business process outsourcing] companies in Madagascar (up from just a handful in 2005, mostly in the capital Antananarivo, employing between 10,000 and 15,000 people (Morocco, the market leader, has 70,000). The reasons companies are flocking to Madagascar is a combination of cost and quality.

With salaries starting at $130 a month (nearly three times the minimum wage) Lalatiana Le Goff, director general at Vivetic, the oldest BPO operator in Madagascar and one of the largest with 1,400 employees, says that Madagascar is 50% cheaper than Morocco with similar levels of quality. “The Malagasies are diligent and have a real desire to learn,” she says. They also have a natural empathy, which, combined with the right training, makes them perfectly suited to handle disgruntled customers.

Then there is the language. The level of French is very good, and customers appreciate Malagasy French. “The tone is softer and slower [than in the Maghreb]; some people have an accent but it’s mild and hard to place,” says Ludovic d’Alançon, chief operating officer at Outsourcia, a Moroccan BPO company that acquired two companies in Madagascar in 2016, which employ 550 people. The time difference is also minimal (one hour in summer, two in winter).

What transformed the sector from data processing niche to stellar digital player however is the arrival of cable internet connection in 2009. Madagascar now boasts the fastest internet speed in Africa (faster even than many developed countries), a pre-requisite for good quality calls and real-time services. Since then, the number of companies has steadily grown.


“They also have a natural empathy”? Is this a weird way of saying “despite where they live, they’re human”? Anyway, a good demonstration of how important internet speed is: Madagascar’s average broadband speed is 24.9MB/s, well above Canada, France and the soon-to-need-plenty-of-foreign-income UK.
link to this extract

Errata, corrigenda and ai no corrida: none notified

7 thoughts on “Start Up No.981: hot v cool v Trump v AOC, AI v Alzheimer’s, the oven that knows what it’s cooking, Firefox kills Flash, and more

  1. re Twitter and Likes: Ars Technica has been trying something old-new with Likes recently, using an optional slashdot-like “Agree”, “Contributes”, “Funny”… list of adjectives instead of just thumbs up/down. I like it a lot because it lets me upvote stuff I disagree with but is still insightful, and takes the guilt out of upvoting silly-funny stuff that doesn’t contribute one jot but is a good chuckle. I’m curious if that feature will be used, improve moderation activity and/or comment filtering use.

    re. Flash: mostly, it isn’t needed any more because HTML does what Flash used to be used for: video/music, animations, interactivity, browser differences neutralization. There are still ways to get Flash… on Android ;-p

    re. Alzheimer’s detection: Isn’t just a matter of humans not being trained properly ?

    re: Intelligent oven: I’m all for it: smart stuff should have either all the smarts or none at all. And I often use the oven example: either it has a “Cook That” button, or it must let me set temp and duration (with physical knobs, not digital buttons). Anything in-between is error-prone, a UI nightmare, and not less work than setting temp and time.

  2. Thumbs-up on the sensor oven too. Anything which offloads human error onto an automated system strikes me as laudable. Sure it’s expensive in the very first implementation which comes out. But this is exactly the sort of improvement which can be expected to dramatically drop in price over time. The first microwave ovens weren’t cheap either, and cooked badly compared to conventional ones.

    Hot take – this achievement isn’t getting the respect it deserves, because home cooking is culturally gendered as woman’s work, so an AI application for it is trivialized (tongue-in-cheek, but who knows, maybe someone will write that as a column).

  3. I’m a big fan of nuclear but the biggest problem you have with rapid deployment is a lack of skilled workers. With certain sites that will remain nameless one of the reasons the costs were supposedly cheap is that they thought they could get away with a the weld quality you get with a submarine, but a containment vessel needs even higher quality than that, so you’re talking something like $2000-3000 per foot to weld compared to $232 per foot for a sub.

    Hence I’m starting to wonder whether the mantra, solar/wind can’t scale quickly enough overlooks that its a heck of a lot cheaper to skill up people to deploy those technologies than in the nuclear industry.

    If I was buying a plant I’d buy it from the South Korean’s, not because they have the best design (they don’t) but because they have the best nuclear engineers at the moment, and its access to engineers that’s the critical component. (the Small Modular Reactor folk will say that this strengthen’s their case but SMR’s aren’t as efficient as a AP1000, and there’s still the long build out time. Smartest thing to do is just build them on the site of existing plants).

    • All good points – the UK industry lost its expertise because of the long delay that the Blair government(s) introduced into building or even commissioning new nuclear. I did think that that WSJ article somewhat glossed over the length of time it takes to actually build out nuclear – you can install a hell of a lot of wind and solar in the time it takes even to get the planning approved for a nuclear plant.

      • Single biggest advancement in energy use would be developing a better grid (for example, if you build an AP1000 you have to upgrade all the grid to cope with the power, an SMR on the otherhand, can slot into the same grid as a coal fired plant because its 100-400 MW compared to 1100 MW+. One of the fascinating developments was that Tesla battery farm in Australia. Paid for itself in about 6 months because its the surges that cause you to overbuild see for other examples

    • I sometimes wonder what’s the impact of sucking up all that energy out of winds and sun. At what stage do wind farms create a micro- or macroclimate ? Does a desert fully shaded by solar panels start to grow grass ? Would even making all our energy that way be unnoticeable in the grand scheme of things ?

      • Minimal in the macro sense. The sun gives us far, far more energy than we could ever use. Catching some energy from the wind doesn’t remove the source of the energy for the wind, which is a combination of the earth’s rotation and temperature differences in different locations.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.