Recommendation engines

Having been on the Internet for much of the past decade and more, it’s been stunning to see how it has evolved. Yes it has, by a lot. The social media platforms, the YouTubes, the other big platforms, have grown and grown and grown into billions of viewers, billions of contributors, and trillions of pieces of content and even more metadata. 

The feel also changed, the feeds changed. It might seem like it was just the mix of people and the mix of content that has changed, but hey perhaps like the american elections and this guy has said, the recommendation engines are partly to blame. 

It’s crazy, isn’t it, that the recommendation algorithms, optimized for innocuous metrics to up engagement, watch time, retention time, repeatability and many more innocent concepts have and seem to promote undesirable material simply because the human mind just seems to love gossip, wild stories, imaginative news and the lure of tabloid content. There must be a reason tabloids and gossip does so well. Hah. It’s dangeous. 

I hope more people are able to distinguish between the truth and the fake, the good and the bad, the real and the unreal, but like our poor sad little human minds, it might be hard to escape the mental clutches of logical fallacies, dreams vs realities. 

Recommendation engines


The rise/rebirth of QR codes is … surprising.

QR code (abbreviated from Quick Response Code) is the trademark for a type of matrix barcode (or two-dimensional barcode) first designed in 1994 for the automotive industry in Japan.” –

The surprising thing is that when the geeks and nerds tried to first introduce QR codes to the world, it was in the form of trying to reduce typing of super long URI paths into browsers and mobile phones. After all, typing is slow and painful when you got h t t p s : / / w w w . e x a m p l e . c o m / r a n d o m / m o r e r a n d o m / m o r e . p h p. Nobody gonna get that right.

But that never really properly took off. Too painful to use, still too much friction.  You never know what you’re scanning. It’s just scary and a bit too slow to scan still.

Then we had link shorteners like bitly, googl etc. Then we had NFC taps. 

But it’s surprising that QR codes came into fashion from a surprisingly unlikely source – China. China and their payment systems. 

What if, this is the death of mastercard and visa, like the deaths of diner and amex. If so much payments bypass them, if people no longer needed credit cards with visa and master on them, if people had new credit cards with other systems (yeah it’s actually different things!), and if people went with less credit, would our banking systems change so much, would shopping data change companies etc. 

What an interesting world.


Can’t Help Falling In Love With You

A heart-aching, heart-wrenching, string-tugger of a song. Surprisingly I never knew it is was an Elvis Presley song, based on Plaisir d’amour [wiki]. 

The other day we heard a superbly slow version of this song, without the background drumset beat. It was, slow. So. heart. achingly. slow. In a good way.

I described it as “when you are hugging someone, in a long hug. and you’re kind of releasing, but the other person keeps on holding you tight. and you’re like ‘oh I guess you want to hug a little longer’. that’s the feeling here when the singer dwells just that bit longer on that note. the very significant small point where it crosses just into that bit of uncomfortably long.”

It’s not easy to describe or understand. But it’s just something you have to know and have felt to understand. 

Not many singers sing it so slow. But it’s quite amazing if you come across one.  YouTube

Can’t Help Falling In Love With You


The World Wide Web / Internet is really actually quite new. Given that I’ve kinda grown up around the growth of the Internet from when it was a new curiosity, to now that it’s entered into all aspects of life and all population. 

It was a fun thing. You wrote a html page, you hosted it somewhere, it worked, people saw it, it was free or low cost, there wasn’t much control, or regulation, or privacy, or security, or discovery. It was just a fun thing that geeky people used to play with, connect with each other, and just use computers for. 

And then it grew. Into more and more useful things that we needed for more and more important things. Health, banking, email, payments, shopping, bookings, etc.

While many of them are extension of normal business models, paid purchases, transactions etc, the core of it, the hosting, the search, has almost always been totally free to the average user. For every paywalled site, there’s been a free open alternative, and it’s difficult to beat free. People will just gravitate towards it – free search, free portals, free social network sites, free email, free videos, free hosting. 

And in that low $0 price, came ads. Ads have been on the internet since forever – stupid old banner ads, silly 1 pixel ad, text ads, picture ads, pop-up ads (near eradicated like measles or something). Ads support so many of the big traditional Internet companies like Geocities, Yahoo, Google, etc. Those who don’t sell hardware or direct services, well it’s really hard to convince people to pay for things online, especially when free stuff works so well. The global state of things also brings challenges in collecting payments, and the intangible software-ness of virtual stuff also is hard to sell, and in tiny bits.

So we’ve had free. And it’s gradually coalesced in a few big companies as things often do, like many corporations and industries, not just online.

People worry whether it is too much, whether they should be broken up. After now the tech giants are really big on paper and financial markets, overlapping the oil&gas giants in market cap and possibly cash holdings.

The single basket risk of so many essential services being provided by a single company also makes it difficult to regulate or ban. 

But I think it will be dramatically problematic if we split them up? All those free ad-supported products will now need to find their own fully sustainable income, their own data streams. If Google were to split up, their top 10 billion user products all need to run their own business models – you would have to pay to use Google Maps, Gmail, YouTube, Google Translate, Google Drive, Android, Google Photos, Chrome, and possibly more. Anyway, either they will all run their own ad engines separate from Google Search, or you would need to pay $5-$100 per month for some of these services, each.

If you hesitate to pay the additional $2 or $10 per month for additional Google Drive / Photos / Gmail storage, what’s going to happen if you need to pay for all these separately?

Not to mention, many backend services like cloud storage, account security etc are shared across all these same-company products. Once split-up, they will need to separately reinvent the wheel going forward or pay someone else for the service. Those free 2FA smses might not be free any longer. 

News agencies have had their own battles against this. Free news, paid news, articles scraped by search engines, it’s thrown the traditional news companies into disarray, and I’m not sure they have gotten out of it.

Frankly the bitcoin and cryptocurrency thing runs in the same vein – decentralized ledgers, lower cost, putting control into different hands (preferably no hands), etc. Disruption of traditional models.

It gets harder when the disruption hits your fundamentals that you kinda need to keep constant or have trust in over your lifetime – wealth, purchasing power, identity, career, etc.

This liberation is like almost like going from a big built-up city in skyscraper and ancillary services, to wild free nomadic liberty in the woods. Free, wild, less constant, more crazy.

What will the world balance look like after the Internet is done with it? 


new* maps and where maps is going

Justin O’Berine – this amazing author who does pretty darn incredible comparisons and analysis of Apple Maps vs Google Maps, has just released another monster blog post with a gazillion comparison images and details on how a new version of Apple Maps released alongside iOS 12 (or something, I don’t track Apple numbers), has evolved differently, especially in California, USA. 

[It’s actually rather odd to think of it as a new version of maps because Apple does things in block releases, while Google is more on this eternal evolving single pane of glass, with ever updating Google Maps data, design and features. After all, it’s a product that doesn’t have any hardware release requirement. ]

So Justin O’Beirne remarked how Google Maps is going down a path where user location history, user current location, user preferences, and all that data is used to predict and customize maps displays, suggested places and many other navigational, directional and suggestion features. This is different from Apple’s approach where they are more on the anonymous user end and being more of a navigational tool. It’s amusing to see how Apple has worked so much on fixing greenery, 3D buildings. Probably looks nice but really how useful is that.

The hilarious part is that Justin pointed out one unique strong point of Google Maps over Apple, and that’s Google’s Local Guides program, where volunteers edit and correct map information all over the world. Millions of active volunteer users who are the best information source about their hometown and where they are at presently. The only other major crowdsourced map info is OpenStreetMaps, but that’s a little different in design. So with this, Google has markedly superior updated local information that is not bought from mapping providers, or algorithmically generated from satellite imagery. It might account for more accurate data points compared to Apple in POI data, and info within buildings etc. Maps data has historically depended majorly on buying map info, using satellite imagery, then using streetview(tm) cars. Now that HERE and Apple both have street cars, what else can they not catch up on? – Local Guides. Would Apple ever be able to get this community going? I highly doubt it, it’s just not the right culture and motivation.

He is right that Maps is becoming less of just a navigational tool and more of a destination in itself to discover places, suggest places, and a bigger community in itself. Almost like a different window or gateway to services. It makes great sense as physical location informs many choices, and after that all of Google’s great scope of business, pricing, photography, search, review, metadata, all come together to provide a great business listing. It’s fundamentally quite different from a Facebook page.

It’s nice to know that someone ought there noticed and attributed a strength of Google Maps to the Local Guides program and the other millions of ongoing or potential contributors out there. The Maps team is right to maintain and support this as there is a huge synergistic symbiosis here – better map data draws more users and features. 

Detractors believe that Google Maps is a proprietary database as compared to OpenStreetMap, but while OSM is a great free resource, it just isn’t as well integrated, convenient and well marketed as Google Maps in the eyes of basic end users. It’s like how you see a “rough” map loaded on your car built-in navigation or shopping mall display, and you just know it’s a less complete or detailed map than the Google Maps you are used to in your daily life. Sometimes the map you are very familiar with, is easier to navigate in.

I would not say Google Maps is perfect – it’s got so many darn problems and is weaker in transit than CityMapper, weaker in restaurant listings than websites/blogs/Yelp/Facebook, weaker in community than Facebook/Instagram, and often not catering enough for local markets. StreetView app is also horrendously slow buggy and tedious at submissions. 

But yeah, I still encourage everyone to start putting reviews into Google Maps, and then maybe photos, and before you know it you’ll find contributing addictive and fun.

Useful links:
follow @googlelocalguides on twitter or instagram
search for #localguides, or #singaporelocalguides, or #letsguide

new* maps and where maps is going

Exhilarating Local Guides Connect 2018

[Cross-posting from Local Guides Connect forum as it’s hard to see stuff in there and it’s a bit more permanent here. Original post]

Exhilarating Local Guides Connect 2018

Oh my, it’s been an exciting week at Local Guides Connect 2018. It’s my second time there, but it’s still super amazing, super exhilarating and really insightful.

Many other attendees have shared numerous joyful photos of the sharing and feedback sessions, vigorous Q&A, deep thoughtful discussions with individual Googlers, outdoor activities to get to know one another, and the incredible number of self organized chillout, partying and supper sessions.

As we look back at all the photos, we are stunned by how much happened within the blink of an eye as the few days sped by, and always wish we could have spoken to more people, attended more events, and taken more selfies.

All the repeat attendees like me try hard to share the same fun experience with the newcomers, organize some extra events to share more of the city, help others to prepare for Connect, and make it more fun for everyone. I hope we have managed to do this. It is part of our gratitude for being able to be here again. It is a family of friends that are a part of our lives now, including literally many families and new babies that have come to say hi.

Many people have also shared on their gratitude and appreciation towards the Googlers – the Local Guides team for organizing Connect and the community in general, and also the Maps product teams for flying out to talk to us (IRL!), share information with us, and taking our feedback. Here’s another +1 from me.

Google Maps and Local Guides are special.

Some things I observed during this year’s Connect made me think deeply about why Google Maps, and Local Guides, are so special.

Firstly, one thing I often hear from the Googlers is how fun the Local Guides Connect event and people are, when in comparison with some other Google events or communities! We are super boisterous and welcoming in all the events, overwhelming in energy, making friends so easily across language and culture differences, sharing stories, language, food, festivals, and spending every single waking minute hanging out with as many people as possible. All the Googlers are also drawn into it as if we are long time friends – you can see photos of many of them at our outside parties.

Secondly, during our small breakout sessions, I realized that fellow Local Guides really care very deeply about our contributions – the reviews, the star ratings, the photos, etc. We want these contributions to be truthful, authentic, and helpful to others. We do this out of an enjoyment in sharing our knowledge with others, and helping make the global maps a more useful tool for everyone. We use and contribute to Google Maps because we realize that we trust the data on it, and we want it to be as accurate as possible. There is an resolute and endearing sense of righteousness and integrity in the community that has never been surfaced or expressed so clearly before. If the Googlers were as surprised as me by this reaction, they sure hid it well.

One world one map

I think these wonderful traits were born out of what we set out to do – helping Google to make the information on maps accessible and useful. We all reach into our Google Map app, seeing the same map as what each other is seeing, the same big database no matter where we are in the world. With a flick of the finger we can be on the other side of the planet, with a search we can find any cafe in the world and see photos of the place. No boundaries or segregation. We cover restaurants, utilities, buildings, shops, markets, bus terminals, religious buildings, hiking trails, gas stations and everything under the sun, crossing all domains from daily living to food, essentials, transport, healthcare.

In Maps and Local Guides, we add our bit of knowledge in, and daily users do not know who has fixed the road or opening hours. Simply it is a better map for everyone and not for a nicer personal profile for us. This culture of sharing, and the interconnectedness of sharing the same map, is why we connect so well with each other. We have always been in this same map, just never face to face before.

It’s the people

After all the (countless) hours in our own phone app, we come to San Francisco and we meet all the other people who has almost unbelievably done the same as us – be equally obsessed about fixing data points and share so many of the same feelings about Maps as we did. They know all the parts of the app, share the frustrations on the same limitations, feel annoyed at getting their edits rejected(!), spend equally long taking photos of places (finally we don’t feel like a crazy person), fix every single place marker, and choose restaurants based on whether it has more than 4.8 stars average. Through this trial of blood, sweat and tears, we still persist.

After two days of making friends, we also suddenly realize that this humble and friendly person beside us has a mind-blowing number of edits, or photo uploads, or video uploads etc. You feel comforted that you’re not crazy for uploading the menu photo of restaurants, and also encouraged that level 10 is possible! None of us worry whether you have level 7, 8, 9 or 10, but my gosh “how did you do that”, and you have a feeling that all these map edits in all countries are done by real life people like you and me. That review describing a delicious milkshake or overpriced burger is written by someone who cared about their experience, and this person is real and standing in front of you.

These new lifelong friends share a strong passion for finding new places to eat, visiting new hiking trails, and travelling to new places. That’s why we are such a fun and immersive group of people, that’s why Local Guides is different from other communities. Connect is fun because of the impressive culture and people, kudos to Google for arranging this. I always tell first time attendees: try to extend your travel and spend more days with the other folks, the golden gate bridge will be still standing next year, but these 150 people will be present together in the same city on the same days for only this one time in your life.

A toast to all you crazy ones out there.

Come for the map, stay for the people. See you all again in streetview* and round the corner.

Local Guides at the top of the world. feat Jens, Pedi and Julien

(*with all the new 360 cameras, I am sure there will be a massive boost in 360 photos, hopefully with not too many of our blurred out hands and faces in it.)

#localguides #letsguide #localguidesconnect #comeforthemapstayforthepeople #allthecrazyones

Exhilarating Local Guides Connect 2018






(女):是否很惊讶 讲不出说话 没错我是说 你想分手吗
曾给你驯服到 就像绵羊 何解会反咬你一下 你知吗

(男):也许该反省 不应再说话 被放弃的我 应有此报吗
如果我曾是个坏牧羊人 能否再让我 试一下 抱一下

(女):回头望 伴你走 从来未曾幸福过
(男):恨太多 没结果 往事重提是折磨
(女):下半生 陪住你 怀疑快乐也不多
(男):被我伤 让你痛

(女):好心一早放开我 重头努力也坎坷 统统不要好过
(男):为何唱着这首歌 为怨恨而分手 问你是否原谅我
(女):若注定有一点苦楚 不如自己亲手割破
(男):回头吧 不要走 不要这样离开我
恨太多 没结果 往事重提是折磨

(女):下半生 陪住你 怀疑快乐也不多
没有心 别再拖 好心一早放开我 重头努力也坎坷
(男):为何唱着这首歌 为怨恨而分手 问你是否原谅我

(女):若勉强也分到不多 不如什么也摔破
(男):好心分手 每天播

(女):难行就无谓再拖 好心一早放开我 重头努力也坎坷
(男):为何唱着这首歌 为怨恨而分手 问你是否原谅我
(女):若注定有一点苦楚 不如自己亲手割破


The IT industry is failing at security

Wow after looking at the huge spate of incidents, it’s quite obvious that the self-touting IT security industry is quite bad at implementing security. It can’t seem to protect against itself? I mean, it’s one thing to blame general dont-know-any-better users, but even the IT industry itself can’t seem to find good, workable, practical, affordable, scalable solutions to its own problems.

  1. usernames and passwords – known to be weak to humans, unable to memorize long digits, unable to keep doing resets
  2. SMS 2FA – was the standard for awhile, and now it’s still being pushed out. However known to have a weak links, unable to protect high value targets. Expensive to scale up. Unable to consistently apply to all logins all interfaces without expensive customization.
  3. Hardware token 2FA – the impractical need to carry a physical token around everywhere, on you, plug into all sorts of foreign computers or personal computers; expensive as hell; even more expensive when you find out you should keep a spare set; fallback to SMS is weak link

So why are we still here? Despite all the brilliant cryptography and mathematics, we’re really not that far along at all. Computers will always be bad at something, there will always be some loophole. We can’t yet have SMS 2FA everywhere and token 2FA is still impractical for most use cases.

Security should Not be too expensive. Usernames and passwords should not be just sufficient. Biometric should not be in the domain of rich and entitled. 2FA should be more built-in.


Could I actually afford 2 x 2FA tokens? S$100? sheesh.

I still can’t get my head around the impracticality of it

  • If I carry a token on my keychain, i need to plug it into the computer to use. I don’t want to carry my keys around everywhere, neither do I want to plug it into a computer – my keys i mean
  • If I carry a token on my staff pass lanyard, i need to plug it into the computer again. what if I forget to take it out?
  • Do I need to take it off my neck? do I pull the extendable wire and plug it in and take it out each use?
  • do we go for bluetooth connections?
The IT industry is failing at security

How the new Apple Watch surprised the world and made Google fans even more jealous.

Admittedly, the ECG function, however unlikely-we-actually-all-need-it, does sound so cool. Like heart rate tracking when it first came out, it is cool, and another order of magnitude higher in difficulty. Lovely. Hope more companies get onto it, and it gets cheaper to make. (However please don’t make all gadgets yet more expensive, ugh. Wallet cries.)

Google fans will lament how the Apple Watch system grows better each iteration, and how Android/Google watches stagnate and grow irrelevant across the years. It’s totally true. I wouldn’t actively recommend Android Wear watches now. Buyers need to make their own choice and live with it, don’t blame me!

It’s a narrow view to see the whole consumer wearable and electronics industry as just these two companies, there’s also the Samsung behemoth, Fitbit, and many others that make lots of phones, watches, wearables, trackers, monitors, etc. Although none are as huge as these two in presence. And many people do purchase Samsung phones (which have some great ones like Note 9), FitBit trackers which have great battery life, etc. Good worthy options.

So let’s assume for a moment the rest dont exist. If our lives depended on Apple vs Google. If the only hope of catching Apple was for Google to buck up, what would the situation be?

phone hardware – great and growing. cpu chips good, display good. innovation on things like screen poor. visual unlock good but lame. security good.
phone OS – so far acceptable. security good. little innovation right now.
watch hardware – pretty good
watch OS – improving steadily in the right direction
tablet hardware – good. steady enhancements.
tablet OS – same as above
voice assistant – poor, and maintaining
maps, backup, search and other accessory software services – poor and maintaining. no big changes to be seen. some investment.

Google (incl Android in general)
phone hardware (assume pixel) – not bad, some improvements. industry standard cpu. not much innovation other than camera which is sufficiently solved for everyone
phone OS – stable, but not much improvement visible. actually lots of underlying work but not sure it’s appreciated. security still needs help.
watch hardware – poor. slow outdated hardware components
watch OS – poor. slow and little support or integration.
tablet hardware – poor and maintaining
tablet OS – same
voice assistant – leader and growing steadily in multiple directions. languages, dual language, functionality, integrations all growing quickly.
maps, backup, search and other accessory software services – leader and maintaining. not much revolutionary innovation lately but small enhancements maintain.

Frankly the two are still, and have always been such different companies. One is hardware based, one is software/data based. If iphones, watches are 80% of Apple, Android and accessories is probably like 20% or less of Google. As a consumer, we see Android and Search etc as the face of Google. Within Google, Search (+ ads) and Cloud are probably the two biggies. I mean, Google Cloud has grown tremendously in terms of revenue, share, focus, manpower. Tremendously. And it has great potential in bottom line, unlike Android which has poor revenue stream other than indirectly through Ads, and which the legislators in EU are giving so much undeserved flak over. If you dont like Android, fine, go and stay within Apple and have no choice and pay through the nose, daft. G as a company does not, would not, and should not, have the focus on expanding/improving Android phones and tablets and watches. It doesn’t bring in sufficient licensing (near 0), nor hardware revenue right now, and foreseeably slow unless the ecosystem business model changes dramatically. Would be an earthquake if it does. Would not be easy or clear path to victory. Might even backfire and cause more problems. Why risk that when they have lots of better revenue on Ads and Cloud side.

If you look at the focus from inside, Google prioritises projects that 1. change humankind (self driving cars, flying cars, quantum computing, AI, etc), 2. are technically challenging and cool (better network, write new OS, containers, cool code), 3. have to do with billion-count audience, free, contains lots of data, analytics, search, India, etc (Mapping, languages). Never forget their core mission of organizing the worlds data and making it accessible – they really repeat it alot and really live it. Do fancy watches and tablets come in here? In a limited way but the important thing is – it’s not the focus.

Companies that don’t focus on something from top to bottom, usually don’t succeed at it. That’s why Google prunes old projects mercilessly. Killing all our favorite apps because it didn’t grow big enough, didn’t fit their plans, or nobody wants to babysit it anymore. Sads.

So we really would not expect Google to give us the next fancy perfect watch or tablet, or phone even. That should and could come from a different company. If we just want an Apple Watch-esque device, Fitbit etc could do the same. It doesn’t need to run 1000 apps with an appstore, just proprietary apps with cool functions would do. My Fitbit charge 2 works great, paired with Android too. (tldr we should give up on Android Wear?) I guess the reason we still root for Google to do this is because 1. we trust Google’s engineering which is usually reputable in software and security, 2. good warranty followthrough, 3. not many companies have the deep pockets to go up against Apple, 4. integrated data would be cool (though so many ppl are anti google monopoly sigh).

So, would the Pixel 3 have anything new? Kind of unlikely to have anything earth shattering. Most software enhancements like Google Lens, AR just don’t have that real-world impact yet. Most new enhancements are sweet but people just don’t market or remember it enough. I love how Pixel 2 has great call filtering (actually all Google dialer), great remote support in app, great Google Assistant integration, great Android P features like smart battery saving etc. But many new Google features are always ported to other phones.

So, we’ll all be watching (ooh pun), but don’t expect any great announcements. But I still strongly recommend the Pixel phones over iPhones. =p


p.s. remember the most stunning part of Google IO 2018? The super smart Google Assistant conversation making remote calls to make restaurant bookings. That, is your indication of where Google is focusing on. Google’s AI vs Apple’s hardware. Each their own unique strengths and obsessions. Dang I want that AI on my phone.

How the new Apple Watch surprised the world and made Google fans even more jealous.