AWS machine learning talk

I enjoyed giving my “Intro to Amazon Machine Learning” talk at the AWS Denver Boulder meetup.   (Shout out to an old friend and colleague who came out to see it.) I didn’t get through the whole pipeline demonstration (I didn’t get a chance to do the batch prediction), but the demo gods were kind and the demo went well.

We also had a good discussion.  A few folks present had used machine learning before, so we talked about where AML made sense (hint, it’s not a fit for every problem).  Also had some good questions about AML, about performance and pricing.  One of the members shared a reinvent anecdote: the AML team looked at all the machine learning used in Amazon and graphed the use cases and solved for the most common ones.

As, usual, I also learned something. OpenRefine is a tool to help you prepare data for machine learning.  And when you change the score cut-off, you need to restart your real-time end point.

The “Intro to Amazon Machine Learning” slides are up on SlideShare, and big thanks to the Meetup organizers.




Running a brown bag lunch series in your office

Courtesy of smoothfluid

Courtesy of maxually

Brown bag lunches are great opportunities for employees to share their knowledge, learn new skills, and bring a small company together.  By ‘brown bag lunch’, I mean an internal presentation lasting about an hour, made by an employee on an interesting topic of their choice.  The name comes from everyone bringing their lunch to work on that day, rather than eating out.

8z has been doing them for over two years, and here are some lessons.

  • Schedule them monthly, and one mont at a time.  Don’t try to schedule out the whole year.
  • Have presenters spend as little time as possible building a powerpoint.  It’s hard to get away from them as a structural crutch, but they don’t really add a lot of value.
  • Bring in real business situations.  One of the most memorable presentations occurred when presenters analyzed a recorded call during the presenation.
  • Have someone be point and recruit people individually.  Don’t count on volunteers, especially at the beginning.
  • It’s OK to miss a month or two if other stuff is going on.  Hello December.
  • Record them if you can.  All you need is an ipad and a youtube account.
  • Technical presentations (like application architecture) are appreciated by the business folks.
  • Everyone has something to say.
  • You can have people repeat every six months or so.
  • Some people won’t want to speak.
  • Presenting in pairs can work.
  • Make sure the presenter leaves plenty of time for Q&A.  8z budgets an hour for the talk and Q&A.
  • Schedule it so founders/executives attend.  This makes a powerful statement and exposes them to direct ideas.
  • Be prepared to capture changes/feedback from the presentation.
  • The departmental cross pollination is a major benefit.
  • Consider themed potlucks (mexican, breakfast for lunch, etc) instead of brown bag lunches.

How do you spread knowledge within your small company?


The future of the web browser…

… from the perspective of people building platforms for add-ons (or plugins or extensions or what-have-you): this collection of videos from the keynote of the Add-on-Con covers a variety of interesting topics regarding browser development, the security model and how extensions fit, ad blockers and what they mean, and more.

Of the major browsers, Chrome, Opera and Firefox are represented–the Microsoft/IE representative was sick, and Apple/Safari was MIA.  It’s about 30 minutes of video, unfortunately split into 5 separate segments.  Well worth a view.

Technorati Tags:


List of Front Range Software Networking Events and Conferences

Updated March 21: crossed out ‘conferences’ because I don’t do a good job of listing those.
Boulder, Colorado, has a great tech scene, that I’ve been a peripheral member of for a while now.  I thought I’d share a few of the places I go to network.  And by “network”, I mean learn about cool new technologies, get a feel for the state of the scene (are companies hiring?  Firing?  What technologies are in high demand?) and chat with interesting people.  All of the events below focus on software, except where noted.

NB: I have not found work through any of these events.  But if I needed work, these communities are the second place I’d look.  (The first place would be my personal network.)

Boulder Denver New Tech Meetup

  • 5 minute presentions.  Two times a month.  Audience varies wildly from hard core developers to marketing folks to graphic designers to upper level execs.  Focus is on new technologies and companies.  Arrive early, because once the presentations start, it’s hard to talk to people.
  • Good for: energy, free food, broad overviews, regular meetings, reminding you of the glory days in 1999.
  • Bad for: diving deep into a subject, expanding your technical knowledge

User groups: Boulder Java Users Group, Boulder Linux Users Group, Rocky Mountain Adobe Users Group, Denver/Boulder Drupal Users Group, Denver Java Users Group others updated 11/12 8:51: added Denver JUG

  • Typically one or two presentations each meeting, for an hour or two.  Tend to focus on a specific technology, as indicated by the names.  Sometimes food is provided.
  • Good for: diving deep into a technology, networking amongst fellow nerds, regular meetings
  • Bad for: anyone not interested in what they’re presenting that night, non technical folks

Meetups (of which BDNT, covered above, is one)

  • There’s a meetup for everything under the sun.  Well, almost.  If you’re looking to focus on a particular subject, consider starting one (not free) or joining one–typically free.
  • Good for: breadth of possibility–you want to talk about Google?  How about SecondLife?
  • Bad for: many are kind of small

Startup Drinks

  • Get together in a bar and mingle. Talk about your startups dreams or realities.
  • Good: have a beer, talk tech–what’s not to like?, takes place after working hours, casual
  • Bad: hard to target who to talk to, intermittent, takes place after working hours.

BarCamp

  • Originally started, I believe, in response to FooCamp, this is an unconference. On Friday attendees get together and assemble an interim conference schedule.  On Saturday, they present, in about an hour or so.  Some slots are group activities (“let’s talk about technology X”) rather than presentations.  Very free form.
  • Good: for meeting people interested in technologies, can be relatively deep introduction to a technology
  • Bad: if you need lots of structure, if you want a goodie bag from a conference, presentations can be uneven in quality, hasn’t been one in a while around here (that I know of)

Ignite

  • Presentations on a variety of topics, some geeky, some not.  Presentations determined by vote.  Presentations are 20 slide and 5 minutes total.  Costs something (~$10).
  • Good: happens in several cities (Denver, Boulder, Fort Collins) so gives you chance to meet folks in your community, presentations tend to be funny, wide range of audience
  • Bad: skim surface of topic, presentation quality can vary significantly, not a lot of time to talk to people as you’re mostly watching presentations

CU Computer Science colloquia

  • Run by the CU CS department, these are technical presentations.  Usually given by a visiting PhD.
  • Good: Good to see what is coming down the pike, deep exposure to topics you might never think about (“Effective and Ubiquitous Access for Blind People”, “Optimal-Rate Routing in Adversarial Networks”)
  • Bad: The ones I’ve been to had no professionals there that I could see, happen during the middle of the work day, deep exposure to topics you might not care about

Jelly

  • Cooperative work environments, hosted at a coffee shop or location.
  • Good: informal, could be plenty of time to talk to peers
  • Bad: not sure I’ve ever heard of one happening on the front range, not that different from going to your local coffee shop

Boulder Open Coffee Club

  • From the website: it “encourage entrepreneurs, developers and investors to organize real-world informal meetups”.  I don’t have enough data to give you good/bad points.

Startup Weekend

  • BarCamp with a focus–build a startup company.  With whoever shows up.
  • Good: focus, interesting people, you know they’re entrepeneurial to give a up a weekend to attend, broad cross section of skills
  • Bad: you give up a weekend to attend

Refresh Denver

  • Another group that leverages meetup.com, these folks are in Denver.  Focus on web developers and designers.  Again, I don’t have enough to give good/bad points.

Except for Ignite, everything above is free or donation-based.  The paid conferences around Colorado that I know about, I’ll cover in a future post.

What am I missing?  I know the list is skewed towards Boulder–I haven’t really been to conferences more than an hours drive from Boulder.

Do you use these events as a chance to network?  Catch up with friends?  Learn about new technologies, processes and companies?


Dan Pink discussed the sad state of employee incentives at TED

Fantastic video from Dan Pink about motivation in the workplace.  He examines the science of motivation, and knocks business for not adapting new methods of encouragement for the new, right brain type problems that face us.

This quote pretty much summarizes the talk:

There is a mismatch between what science knows and what business does. And what worries me, as we stand here in the rubble of the economic collapse,  is that too many organizations are making their decisions,  their policies about talent and people, based on assumptions that are outdated, unexamined, and rooted more in folklore than in science.

You can see a transcript of the talk by clicking on the ‘Transcript’ link on the right hand side of the video.  It’s actually pretty cool–clicking on a sentence and it updates the video to that part.  It’s not linkable, though–FAIL.

A few takeaways:

  • If/then rewards work well for simple tasks…they concentrate the mind and narrow your focus.
  • From a study by the FRB of Boston, “once [a] task called for ‘even rudimentary cognitive skill’ a larger reward ‘led to poorer performance'”
  • Management is not a tree, it’s a television set.  We invented it.
  • Atlassian used to give their engineers autonomy at least a few times a year to choose what they work on (FedEx Days), and now gives workers control over 20% of their time.  (Joel has something to say about Atlassian too.)
  • Autonomy, mastery, purpose are what people are looking for (once money is taken care of)
  • Results only work environment–people can work when they want, as long as they get their work done.  Here’s a stub wikipedia article about ROWE.
  • Encarta vs Wikipedia–who won?  The encyclopedia that leveraged people’s desires to work, not the one that paid them.

Now, my thoughts.

  • First, watch the whole video.  It’s only 18 minutes and is well worth your time if you are an employer or an employee (which covers most of us, I think).
  • ROWE reminds me a lot of college, especially higher level classes.  No one cares about when you do the work and I don’t remember being required to be at classes, but the results (passing a test, turning in a paper) were very important.
  • He only talks about the autonomy component of the new ‘motivation trilogy’ (autonomy, mastery, purpose).  I wish he’d chosen to talk about ‘purpose’ because to me that is the hardest bit–someone needs to do grungy jobs.  I guess granting workers autonomy is pretty revolutionary too.
  • “once money is taken care of” is a huge elephant in the room that he again does not address.  When the work is high value add, it makes sense to take money off the table (software developers are very lucky in this respect).  But what about a worker at Target, for example?  A Target store can’t afford to pay someone enough to take the money issue off the table, but can probably benefit from the ‘motivation trilogy’
  • The whole Encarta vs Wikipedia example that he gives is great, but he ignores the fact that Wikipedia has very few paid employees and that Wikipedia only won because of volunteer labor.  It’s not a solution that scales across a society.

Overall, in general, a thought provoking talk (expect nothing less from TED).  I would say that he is describing the future of work as consulting.  You are paid for what you know, the problems are fuzzy, answers are unexpected and at times unclear, and results arrived at matter far more than hours put in.

Is the future of work consulting?  If so, the business world is about to be upended, because, to borrow Dan’s phrase, “the operating system of business” isn’t designed to handle a workforce of consultants.  Heck, society isn’t either.

Technorati Tags: , ,


Interesting GWO Case Study

I’ve written before about Google Website Optimizer.  But it’s always nice to see hard data.

Here’s an interesting GWO Case Study I found online, via a presentation by Angie Pascale.  It focuses on optimizing landing pages for a college system.  Conclusions:

Although the SEM agency did not find a correlation between brain lateralization and form location, they did succeed in optimizing Westwood’s program landing pages. On average, the program pages saw a 39.87% conversion rate improvement, with 83.1% being the highest upgrade. After significant results were revealed, the agency stopped each experiment and changed the format for every page to reflect the best-performing contact form location.

Technorati Tags: ,


“Blending Social Media with Traditional Marketing” Presentation

I have some interest in social media–I obviously blog, but I also have a twitter account as well as some other ‘social media’ interaction.  It’s also the reason I attend the Boulder New Media Breakfast.  One of the ways I keep up to date on topics of interest is using Google Alerts (it’s not just for Craigslist).  I also use FiltrBox, mostly because I have some friends who work there.

The FiltrBox folks are having a 1 hour webinar tomorrow titled “Blending Social Media with Traditional Marketing” which looks to be pretty interesting.  From the description, they’ll discuss:

how social media marketing and traditional marketing integrate, how your company can leverage social media, along with best practices, how to listen, monitor, engage and interact, with highlights of specific case studies.

Not sure how much marketing material you might get from signing up, but you can always use a throwaway email address.


Denver BDNT Sep 2009

I helped Brian Timoney, of the Timoney Group present last night at the Boulder Denver New Tech Meetup.  It was my second experience presenting at BDNT.  (I presented in Jan of 2008 on GWT.)  But it was my first time at BDNT Denver–down at the Tivoli.

Co-presenting is always different than presenting alone.  I actually had a pretty small role in the presentation–I mostly just drove the demo (underwater navigation with Google Earth to visualize sonar coverage data–it’s very cool, but I don’t feel comfortable putting the demo login up–contact me if you want to see it).  I worked with Brian on the presentation format.  Brian has deep knowledge of GIS concepts (he recently ran a workshop at GIS In the Rockies), but he’s used to having more time to cover concepts, and 5 minutes just enforces a certain brevity.

We had a mentor–Josh Fraser of EventVue took some time to run through our presentation with us.  It was really great to have a third party, especially one in tune with the BDNT, give us feedback.  As I told Robert Reich last night, we went into the mentoring session with one presentation, and left with an entirely different one.  If you’re thinking about presenting at BDNT, please get a mentor (and you might have to ping the organizers a few times to get one–we did).

If I ever present at BDNT again, I’ll follow the format we arrived at:

  • 15 sec intro
  • 1 min talking about problem
  • 2-3 min demoing software solving problem
  • wrap up
  • contact info on screen during questions

However, one of the difficulties in presenting for 5 minutes to a varied audience is that it is hard to know what knowledge to assume (about, say, GIS).  I talked to some people after the presentation, and it seemed like we assumed our exposition of the problem was better than it actually was.  I guess one way to address that would be to have a 30 sec intro spiel that you could deliver or not deliver based on a show of hands.  Not sure if there are other ways to deal with this issue.

Finally, we were the only formal presentation last night.  It sounds like BDNT Denver isn’t as supported by the community as BDNT Boulder, in terms of participation.  I hope it doesn’t end–so, if you’re in Denver, consider attending this meetup–it’s a great place to network and get excited about tech.  Here’s the calendar of meetups.

Instead of other presentations, we went unconference style, a la BarCamp.  People broke into 5 groups and discussed a tech issue (personalization, structured data, real time web) in detail for 10-15 minutes.  Then someone from each group presented 1-3 minutes.  The twitter feedback seemed pretty favorable.  I like BarCamp formats, and enjoyed the change.  I found that everyone in my group had lots to say about personalization, including some pretty creepy personal storied about advertising on the net.  I believe someone was going to write up the resulting presentations–will link to it when I find it.

Technorati Tags: , , ,



© Moore Consulting, 2003-2017 +