Not sure if this is the one Shirley Bassey and David Bowie used. Photo by avlxyz on Flickr.
You can now 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 10 links for you. Use them wisely. I’m charlesarthur on Twitter. Observations and links welcome.
How reliable is my train? It’s something I often wonder when choosing which train to take, and it’s something that’s hard to answer without months of commuting experience.
This summer, in partnership with The Open Data Institute, we built a web site to try and help you find an answer to exactly this question. So if you’re are regular train traveller, or just a bit of a stats geek, why not head over the the Fasteroute Delay Explorer and plug in the details of your journey, and see how close to timetable your train generally runs. You’ll also be able to see how it compares to other trains around the same time. You might even find an excuse to stay in bed a bit longer in the morning.
Windows 10 had a good holiday.
That’s the prognosis from Microsoft, which reports that the latest version of Windows is now on 200 million devices. A good chunk of those were activated over the holiday season — 40% of Windows 10 devices became active on Black Friday or later.
Windows 10’s adoption is faster than an previous version of Windows, according to Microsoft, outpacing Windows 7 by 140% and leaving Windows 8 in the dust by a massive 400%. Microsoft also reports that 76% of its enterprise customers are in “active pilots” of Windows 10, but it’s hard to know how relevant that is without knowing the scale of those pilots; some companies may be testing a relatively small portion of their PC footprint, for any number of reasons.
Still, the numbers are all pointing in the right direction for Microsoft’s goal of having more than 1 billion devices running the OS within two to three years.
Here’s an idea: what if Sean Penn’s visit and subsequent terrible Rolling Stone interview with Mexican drug lord El Chapo was actually set up by the DEA/FBI/CIA/spies in order that they could nail said Chapo? I know, you laugh, until you read this by Marcy Wheeler, which lines all the ducks up, including this:
Perhaps the most interesting detail is that when Chapo asked Penn to come back in 8 days for a return visit that never took place, Penn responded by asking for a photo — for Rolling Stone. Except that he arranged it so that it would be usable for facial recognition.
I say I can. I ask to take a photograph together so that I could verify to my editors at Rolling Stone that the planned meeting had taken place.
I explain that, for authentication purposes, it would be best if we are shaking hands, looking into the camera, but not smiling. He obliges. The picture is taken on Alfredo’s cellphone. It would be sent to me at a later date.
Who knows? Maybe Rolling Stone uses sophisticated facial recognition software in the wake of their University of Virginia rape story disaster?
Oh yeah, also: pretty much immediately after Penn’s visit – set up via much cloak and dagger – El Chapo came under siege from Mexican troops. Pure coincidence, I bet. Totally. Sure.
link to this extract
A word about Shirley Bassey. During the very early days of Ziggy Stardust, we often used to play these fairly grotty clubs called the “workingman’s clubs.” They were sort of like nightclubs but you got a cheap meal. The whole family would come. A round of beer. A rock act. A stripper — sometimes one and the same. Well, backstage one night I was desperate to use the bathroom. I was dressed in my full, battle finery of Tokyo-spaceboy and a pair of shoes high enough that it induced nose bleeds. I went up to the promoter — actually I tottered over to the promoter — and I asked, “Could you please tell me where the lavatory is?”
And he said, “Yeah, look down that corridor. On the far end of that wall. You see that sink? There you go.”
I said, “My good man, I’m not taking a piss in the sink.”
He said, “Listen son, if it’s good enough for Shirley Bassey, it’s good enough for you.”
From which I learned that mixing elements of bad taste with good would often produce the most interesting results.
I will throw Paul Graham a bone for recognizing that in terms of scale and impact on the American economy, Wall Street is definitely the bigger concern.
But my guess is that what probably infuriates you about Paul Graham’s essay is his tacit contention that startups create wealth.
This is not true.
First of all, over 95% of startups fail. Every venture capitalist knows this. Those pesky things, for the most part, just eat money and more often than not actually destroy wealth.
But the second reason why you should not allow yourself to think that startups create wealth is because of how they are funded.
What actually happens is wealthy people like Paul Graham fund startups because they think these things are valuable. Through venture funding, rich people legitimate startups. Thus, they confer value upon the startup. They then use their ridiculous money and connections to “advise” and “mentor” those they deemed worthy of capital so that they can use this capital to build a future people like Paul Graham expect to see.
What Paul Graham never dissects in his essay is that people like Paul Graham simply take it for granted that they’ll be the ones to decide where capital goes.
I’d like to examine those numbers in Graham’s essay about big and small companies back in the 1950s/60s too. But this is a great – worthwhile – read once you allow yourself to consider it with an open mind.
link to this extract
Rebooting Android’s 10 millisecond problem: audio latency improvements in Android 6.0 Marshmallow » Superpowered
Since we published “Android’s 10 Millisecond Problem: The Android Audio Path Latency Explainer” in early 2015, Google has made tremendous strides in improving round-trip audio latency on Android OS. With the deployment of Android 6.0 Marshmallow, the Huawei Nexus 6P clocks in at a much improved 18ms round-trip audio latency and the HTC Nexus 9 at 15 ms.
As readers may recall, 10ms round-trip audio latency is the threshold that must be met by Android to be considered truly ‘pro audio’.
One second of round-trip audio latency sounds like an eternity. 250 ms is still terrible. 100 ms is 10% of one second, and still sounds unacceptable. At 25ms, things begin to approach acceptable ranges. Unless you are Terence Fletcher, the nightmare-ish jazz instructor from the movie Whiplash, 10 ms should be perceived as instantaneous…
…In our previous article, we discussed that the the Android Audio Hardware Abstraction Layer (“HAL”) implementations, the component between the audio driver and the media server, are often poorly implemented across the the Android device landscape. Google has also ensured that the HAL has been implemented properly for the most recent Nexus devices.
However, the Android media server itself does not look like it has been significantly improved from Android Lollipop to Android Marshmallow. While it was already good in Lollipop, it appears as if Google is now hitting the hard limits of the media server’s “push” mechanism.
Seems like pro audio latency is, once more, something that Android will have next year. Meanwhile, the listing of device latencies still shows all iOS devices at below or only just above 10ms.
link to this extract
it’s hard to tell what’s going on inside an effective neural net, and even harder to understand in what ways they work like real brains and in what ways they do not. But now that we know they do work, we need to know how, so as to improve the next generation.
That’s the utility of the Deep Dreams process. For instance, in one kind of experiment the researchers would choose which layer of the net would be active to enhance patterns it detected in a random photograph. If they chose one of the lower layers — those making the system’s initial assumptions about what an image contains — they would get intricate patterns, because at that point the network is analyzing the edges of objects and not yet classifying them. Another type of experiment tapped the higher layers, encouraging the system to riff on what it had begun to recognize. That’s when the weird animals will appear. While the output is fascinating, we’ve learned more about the way neural networks operate.
But [Alexander] Mordvintsev’s experiment is important in another way: as a pointer to the vast potential of neural nets. As these nets develop, they are destined to not only match human ability in some areas, but exceed it.
One of the notable quotes comes from a
Swedish Swiss artist who argues that in five years we’ll be using this sort of thing inside – or instead of? – Photoshop.
link to this extract
The big push by utilities across Australia to hit solar households with higher network charges is underpinned by the claim that rooftop solar does little to reduce peak demand.
There is increasing evidence that that is not the case. Peak demand has been pushed in some states to the evening, after the sun comes down, but what is often not displayed is what the peak would have looked like without rooftop solar.
In short, it would have occurred earlier in the day, and at a much higher peak. This is critical, because networks super-sized their grid in anticipation of big rises in peak demand. The combination of energy efficiency and rooftop solar and declining industrial demand has junked those forecasts. But we’re still paying for the investment.
This graph released last week by the Australian Energy Market Operator, in a presentation on the WA market that it now manages, illustrates the point in Western Australia.
The peak – without solar PV – would have appeared at 3pm in Perth, and be considerably higher than the peak level with solar PV, which now occurs at 4.30p. Yet still, the network wants solar households to be hit with higher network fees, another example of where the benefits of rooftop solar are not factored in.
Gotta love big business blaming people for doing the right thing, and finding a way to make it seem like it was the wrong thing.
link to this extract
In response to requests from publishers, Apple said it now allows the integration of features from measurement specialist comScore into the app. Apple initially provided updates about the app’s usage by emailing spreadsheets that give a high-level overview of usage, but publishers want a more robust self-service dashboard to access that information.
Media outlets say they have questions about advertising, too.
If they sell their own ads, Apple requires that publishers use its iAd advertising technology to insert, or “serve,” them. Many publishers would prefer to use ad-serving tools provided by other companies such as Google. Requiring publishers to devote resources to a tool they only would use for Apple News could make them question whether it is worth it.
Mr. Cue said he was surprised by the extent to which publishers call on Apple to handle ad sales. He said Apple has accelerated the development of its iAd network and expects to launch a self-service ad-buying platform in the next two months to help increase ad spending.
Thin end of a wedge which Apple might want to consider. As surely as night follows day, advertisers will want to retarget ads across different publishers within Apple News, which means personalisation and user identification. Is news really that important?
link to this extract
Instagram has about 300 million active users and is projected to make about $700m this year. One might then assert that $2 a user is a reasonable target for a social media app that is light on ads. Heck, I honestly haven’t seen an ad on Instagram outside of screenshots in news stories about ads on Instagram. Reddit is a popular social media site that has about half the users of Instagram with about 160 million active users. How much do you then think they made last year? $350m? $175m? $100m? $80m?
Nope. The answer is $8m. That’s 5 cents a user.
What Reddit has found out the hard way is that their advertising doesn’t fit natively into their platform.Their ads often don’t match the form of the content and when it does, it doesn’t match user intent for what they want out of Reddit. On the other hand, people go to Instagram to see beautiful photos. Beautiful photos from brands they’ve expressed an affinity with via Facebook or Instagram’s social graph are the epitome of a native advertising slam dunk. The results advertisers have seen speak for themselves.
Is Snapchat like Reddit or like Facebook? Snapchat’s original product is actually quite bad from an advertising perspective. When you launch it to send messages you start directly in the camera so no place for ads. Secondly, ads into the user’s inbox of received messages or as part of message viewing would be extremely disliked by users and isn’t aligned with user intent.
He has a point. Instagram’s ads (“sponsored posts”) are pretty dire, too.
link to this extract
Errata, corrigenda and ai no corrida: none notified.