Google Photos Logo

How Google’s AI can eliminate the need to keyword photos

Here’s a video I made that points the way forward when it comes to indexing and searching our photography libraries:

Google has built some truly amazing object and scene recognition AI (artificial intelligence) into its Google Photos service. By comparison, the drudgery of manual keywording we currently have to do, not only after we import our photos into the software we use to manage them (Photos, Bridge, Lightroom, etc.), but also when we upload our photos to stock photography websites (for those of us who do that) is downright archaic. The artificial intelligence algorithms that Google uses on the photos uploaded to its service do all of that automatically. They index every photo and identify every object, scene, face, logo, etc., whatever you can think of, and they’re constantly being improved.

You don’t have to be a pro photographer to take advantage of this AI. Even your personal photographs become easily searchable once you upload them to Google Photos, without any manual keywording. Try it and you’ll be amazed, just like I was.

By the way, Google is not paying me to say this. I just love the work they did on their photos AI. Furthermore, I wrote a post critiquing their buggy desktop Backup and Sync software which uploads our photos to their platform.


My Drobo review is first at Google

A couple of days ago, I noticed an increase in the traffic to my Firewire Drobo review, most of it from search engines, so I did a quick search on Google for the phrase “drobo review“, which is what people were using to find me. To my surprise, my review was the first search result that came up! I’d been in the #2 spot for a long time, just under CNET, for the same phrase, but now, without having made any changes to my review since I’d written it, I ranked first.


This makes me happy, because when I created my site, I wanted to sit down and write good articles while staying away from any unethical SEO tricks or even white-hat SEO tricks like keyword loading and other such unappealing, tedious stuff. I just wanted to create good content and get noticed because of that, not because I’d tricked the search engines into ranking me higher up the page. That would have been an empty success indeed.

It also makes me happy because I like my Drobos. So far, they’ve worked well for me, and I’m glad I’ve found a reliable and expandable way to store all my data. It’s also worthwhile to note that my Firewire Drobo review was published months after it came out officially. I did not get a review unit, I didn’t have to pull any strings to be among the first to get one, and I didn’t spend a feverish night working on my review after it first came out. You know how the press clamors to get review units of products when they first come out… I didn’t do that, and it’s very refreshing to see that after taking my time and really putting my Firewire Drobo through its paces, intensively, for a prolonged period of time, I was able to write a truthful review that is now ranked first at Google.

It’s been about three years of intensive writing, and my work has begun to pay off. (I began publishing multiple articles per week in 2006. I’d only been publishing sporadically until then.) In 2007, almost two years ago, I noticed I was getting more and more traffic from search engines, and made a list of the articles that were getting noticed. For a lot of them, I was either on the first page of search results, or among the first few search results, right at the top.

Still, it’s something to be the first search result for what is a fairly common tech phrase such as “drobo review”, and it really makes my day that I, a writer working alone, using WordPress and hosting my site on my own little Ubuntu web server at SliceHost, has outranked CNET and other big names such as Engadget and others, on Google, the world’s biggest search engine. It serves to illustrate very well a point Matt Cutts from Google has made time and time again: just focus on writing good content, and the rest will come. You’ll get indexed, and as your site builds a larger collection of articles, your online trust will cause you to rise up among the search results, until you make it to the top. You don’t need tricks, you don’t need to get headaches from trying to squeeze SEO juice out of every paragraph and page title and others — you just need to write informative articles.

I’d like to thank God for this. You see, I live by certain principles which are rooted in my religious beliefs, most notably in the Ten Commandments found in the Bible. When I began to write online and created my site, I didn’t want to steal, and I didn’t want to lie. Taking content from others (content-scraping) is theft, so I don’t condone it or do it. Using dirty SEO tricks to rank higher in search results is also theft, because those who do it are robbing others of those spots and robbing tech engineers at search companies of their time, which they will have to use to modify algorithms and clean up the search results. And using those same dirty SEO tricks is effectively a lie, because those who do it are misrepresenting their websites and their articles. That’s not me, I don’t want to do those things, and I’m really glad to see that God proved me right when I stuck by my principles. I’m also glad to see that a company such as Google exists, and that it rewards honest, forthright behavior.


Photographs for sale

My photographs (a select number of them) are now up for sale through both Alamy and the PhotoShelter Collection. Here are the links:

Alamy presents you with a search page first. To see my photos without doing a search, just click on my name. PhotoShelter presents you with the photos right away.

Of course, you can still purchase any of the photographs I post here by simply contacting me via email and indicating the photo you’d like to purchase. More info on this is available on my Photos for Sale page. But for those of you that prefer a more streamlined look and an instant price quote, Alamy and the PhotoShelter Collection should do nicely.

If you should lose the links to my sites at Alamy and PhotoShelter, don’t worry, I’ve made it easy by listing them in the sidebar. Just look for this section:

Purchase my photographs

Happy shopping and many thanks!

How To

How to hack Windows Desktop Search and turn it into a dictionary

If there are some of you out there using the beta version of Windows Desktop Search – you know, the one that imitates both Spotlight and Google Desktop Search – then you probably know it’s alright.

I’m amazed that Microsoft has actually turned out a usable piece of software that doesn’t crash. I like it because it starts finding documents as soon as you start typing (like Spotlight), and the search index is live, unlike the Google Desktop Search, which is time-delayed, and also contains bad entries for files you’ve moved or deleted.

When you search with WDS, you get the option to search your Desktop, or the Web. Well, there’s one huge caveat: don’t bother searching the web with the Microsoft Live Web Search, or MSN Web Search, or whatever the heck they call it these days. It’s downright pathetic, and if you’re using IE 7 Beta, it may even crash the browser. Fortunately, Google has come up with a solution. Just surf over to their main site, and within IE 7, you’ll get a JavaScript note (not window) in the top right corner of the web page, that will invite you to set your IE search engine to Google’s. Run the tool, and do that.

Now, for the fun dictionary hack. In the Windows Desktop Search bar (located in the taskbar), type the usual Google shortcut for searching definitions: “define:word”, then click on the Web button. Bingo! Now you’ve got your very own dictionary, neatly accessible right from the taskbar. Enjoy!


IconSurf: the visual search engine

I’ve been meaning to post about IconSurf for some time. It’s a cool site/search engine that lets you search for other sites based on their favicon graphics – you know, the .ico files that reside at the root level of sites, they show up in your browser’s address bar, next to the URI. What’s stopped me is that the site loads very slowly. I thought its author would do something about it, but since nothing’s changed in a few months, I guess that’s about as fast as it’ll get.

Still, as frustrating as it is to use, it’s an interesting concept. You can alphabetize the sites in its directory (over 40,000 at the time of this writing) by TLD (top-level domain), by starting character (a, b, c, d, etc.) or search by a keyword or phrase within the URI. Obviously, this is a basic sort of search functionality, but it does let you narrow your search, and as I said, the site is interesting-enough overall to make you forget about its limitations.

Just think about the possibilities! If a proper image-search engine is put in place, you could organize the sites by color schemes (primary colors, pastels, etc.), by shapes (round, square, oval, etc.), by colors (ex: find a site that uses red in its favicon). This engine has possibilities!

I encourage you to give it a try! Go to it, then step away from the computer for a couple of minutes while it loads. When you come back, you’ll have a nice surprise. Enjoy!

How To

Moving email from a PC to the Mac


For the purposes of this guide, I’m going to assume that you were using Outlook or Outlook Express on the PC, and you want to import your messages into the Mail app that ships with Mac OS X. I speak from personal experience with this guide. I had to do this when I switched from a PC to a Mac in September of ’05. I’ve finally solved the last piece of the puzzle, and I wanted to share this information with you so it won’t take you as long as it took me.

I had an email archive that spanned close to 10 years – a fairly complete one at that. I’d been keeping it in Outlook Express, then Outlook, over the years. When I switched to the Mac, I didn’t want to lose it. My wife had an archive that spanned a few years as well, and she kept it in Outlook. We didn’t want to lose those messages, either. As you know, there’s no easy way to import email directly from the PC into the Mac. There’s no nice and easy transfer wizard, for multiple reasons: different operating systems, different file systems, different ways of storing the mail, different applications, etc. Even if there had been a wizard, given my huge mail archive, I’d have probably crashed it.

What didn’t work

Still, I tried to be reasonable. I thought equivalent Microsoft products on the two platforms might be able to import from each other. So I took my PST from my PC, transferred to my Mac, and opened Entourage (the Outlook equivalent on the Mac). The import failed. The two can’t talk to each other. You can’t import between them either way. I thought that was pretty silly. Of all the things Microsoft does wrong, this has to be one of the more obvious ones.

Next I tried the less possible, which was to import from the PST file directly into Mail. That didn’t work, either. I surfed the internet for solutions, and stumbled across a possibility of installing Eudora on the PC, importing from Outlook into it, then copying the library onto the Mac, and using some special utility to do the PC to Mac translation. Well, Eudora failed on the import from Outlook. Again, I had a big PST, I wasn’t surprised. Plus, even when I tried transferring the messages it had managed to copy to the Mac, the utility didn’t do its job. At any rate, I hadn’t put my hopes in Eudora. It might have been all the rage in the early nineties, but it’s pretty useless now. Somebody else suggested using old versions of Netscape Mail. I tried that as well, only to fail again.

I called Apple Support, who were completely clueless about this. Finally, on my second try, the technician suggested I use an IMAP account to transfer the email between the two computers. I saw two problems with that, both related to the size of my archive: one, where am I going to find an IMAP account with more than 4GB of storage, and two, I’m not going to sit there and upload over 4GB of data through my DSL connection. It was going to take days, if not more. Obviously, not a very practical solution.

Updated 1/3/08: Gmail now offers more than 6 GB (and growing) of email storage, and includes both POP and IMAP access.

What worked

Just so I won’t drag this out needlessly, Thunderbird turned out to be the best solution for the transfer. I installed it on the PC, imported from Outlook into it, then transferred the mailbox files to the Mac, where I had to delete the mailbox index files (.msf) files, and only leave the un-indexed data files there. That’s because the Mac version of Thunderbird needs to build its own indexes. So, I located the directory where the email got stored for Thunderbird on my Mac, moved the mailboxes there, and deleted the index files. I then opened up Thunderbird, and after it re-built the indexes, there were my messages! After all the trouble, I was pretty happy!

Next, I wanted to get my email into Mail. This is the step that took the longest for me, and I just solved it yesterday. Granted, I hadn’t been looking very hard since last September… Now, some of you might be asking yourselves why in the world I’d want to switch from Thunderbird to Mail, and I’ve got two reasons: one, and the most important, Spotlight indexes Mail messages, so I can search for what I need from one location, and two, iPhoto sends out photos through Mail, and we email photos a lot; we wanted to have our email messages in one place. Yes, I know, we should share our photos on the web instead, etc…

So, how did I solve it? Certainly not by calling Apple Support, who are were clueless on this issue as well. And I also didn’t solve it by importing from Thunderbird into Mail, which is impossible (not any more), as you might find out if you try it. Version 2.0.7 of Mail crashes miserably, and has done so, reliably, since September of ’05, on my every attempt to import from Thunderbird. I choose Import, then I select the Other or the Netscape/Mozilla option (since Thunderbird isn’t listed as one of the options), then I browse for the location of the message databases, and when I select Import, it crashes like a drunk limousine driver.

Instead, I solved this by doing a search on the Apple Support forums, where intrepid users have posted some great solutions. Among them, I found the Eudora Mailbox Cleaner. It’s a wonderful little utility that will let you drag your Thunderbird message databases onto its icon, and will automatically convert them to Mail message databases. It will also copy them to the proper Mail message library. All you need to do is sit back and wait for it to finish, then rebuild your Mail folders, and all your messages will appear – just follow the directions you’ll find on their website. The best part is that it didn’t crash while it processed my entire archive (over 4GB of messages)! Now that’s a reliable application!

Updated 1/3/08: It turns out, as one of the commenters has pointed out, that Leopard’s version of Mail includes an import function from Thunderbird. Problem solved. Thanks Logan! Now I wonder if a new version of the Eudora Mailbox Cleaner will be released, or whether this new import feature in Mail will negate the need for it.

When I got done, I was ecstatic. All my mail is indexed with Spotlight, and I can instantly find messages and files that are years old without having to do separate searches for each!

Let’s review

  1. Install Thunderbird on PC, import from Outlook/Outlook Express into it.
  2. Install Thunderbird on Mac, note storage location for mail files.
  3. Copy message databases onto Mac, in the specific directory where mail is stored, delete index files, then start up Thunderbird and let it rebuild the indexes.
  4. Use the Eudora Mailbox Cleaner to export to Mail. (only for OS X Tiger)
  5. Rebuild mailboxes in Mail, then relax, because you’re done! (only for OS X Tiger)
  6. If you have Leopard, skip steps 4 and 5, and use Mail’s import function to get your messages out of Thunderbird. (only for OS X Leopard)

💡 Thunderbird and the Eudora Mailbox Cleaner are free software. If you find them helpful, please don’t forget to donate to them, even if it’s just a few dollars. It’s the right thing to do if you want to support the efforts of their developers. Here is a donation link for Mozilla (the maker of Thunderbird), and here is a donation link for the Eudora Mailbox Cleaner.


The story of Comeacross and Doublecross

I did a search for ComeAcross on Google today, and stumbled upon an old folktale printed in Bagdad in the early years of the 20th century. It’s recorded on the website of Allan D. Corre of the University of Wisconsin at Milwaukee, where it’s part of a collection of popular Arabic literature of the Jews. Read it, it’s nice, and the moral is good. I’m glad to see Comeacross behaved wisely even before it became an informative website! 😀

Updated 2/17/08: ComeAcross was the name of my site from May 2006 to December 2007.


A review of SimplyHired

SimplyHiredI’ve been using SimplyHired for the past few days, and wanted to share my impressions with you. The site is an aggregator. It pulls job listings from all of the major job sites and filters them for you based on your search criteria. At my latest check, they were indexing over 5,381,261 jobs. That’s a LOT of jobs!

I love the search by zipcode feature. I was able to punch that in, plus a keyword, and boom, I got a listing of all the jobs matching that keyword near my home. Beautiful! What’s also very cool is that the site’s using AJAX, so when you save something, the entire page doesn’t have to refresh.

Without creating an account, I was able to save jobs to look at them later. When I decided to create an account, those same jobs I’d saved showed up under it after I logged on. Nice!

You can rate jobs and put comments next to each, like, “applied 4/24/06”, which is what I did. (Incidentally, I’m looking for a job…) The jobs are then sorted automatically based on your rating, with the highest-rated ones bubbling up to the top.

The design is beautiful and clean. The color scheme is great. The functionality works better than advertised. What can I say, I love it!


Google's live tests

Just stumbled onto another entry on the Google Blog which helps explain why I’ve been seeing reports on other blogs of different interfaces for Google’s Search or other products. Apparently they choose to use (at random) sets of live users, and open the additional functionality to them for a limited time, to see how they’d interact with it. The reason is simple: there’s no substitute for real-world testing. This is pretty cool. See the entry here.

How To

A poster of Google search commands

From the Search Engine Watch Blog: “Can’t remember special commands that help in doing special searches at Google (and often work elsewhere, as well). The latest edition of Google’s newsletter for librarians points to two posters you can print with the commands. Suitable for framing — well, for tacking to a wall, you might find them handy. Need something more comprehensive? There’s also the long-standing Google Cheat Sheet…” Here is the link.