14. September 2010 01:12
I know someone that works in social computing - the domain of Twitter, that hates the app. Any discussion of Twitter is met with a growl of disgust.
Jaron Lanier's recent book, You Are Not a Gadget: A Manifesto, slams the culture of Twitter feeds and Hive Mind thought.
I don't hate Twitter, but I might use it a little different than other users. Twitter is essentially just a scratch pad. If I am in a meeting and hear something interesting, I might post something on Twitter that I can reference later. At conferences I post to Twitter instead of jotting down notes. This gives me a quick reference any time I want to look back at the conference. I've found it easier to use TweetDeck than a notebook, and I reference it more.
About once a month I check in on my Twitter feed. It gets very boring very quickly. Most of the posts just regurgitate some marketing post or give a shout out to the latest repackaging of the same trend metrics. Look at the posts about Android - millions of posts about Android passing iPhone, very few applying any kind of critical thought to what it means.
So there you go - if you want to look over my shoulder at my scratchpad, its Twitter. I am really only talking to myself, something I wish more people would do.
1. June 2010 01:18
I recently had a chance to participate in Andrew McAfee's panel on Enterprise 3.0. Since I am still wrapping my head around Enterprise 2.0, the concept was a bit of a stretch for me. Still, it was great to hear the ideas of everyone on the panel and some of the folks in the audience (MIT CIO Symposium). Before I go back to my day job I thought I would jot down my thoughts on the subject. There's one point in particular I am still noodling on.
Dr Edward Curry made a fantastic case for connecting stovepipes of data. His assertion was that distilling meaning from the stovepipes will be the key enabler for Enterprise 3.0. Linked data technology will change the way businesses interact with data. He did a great job differentiating between documents (2.0) and data (3.0).
The highlight of the panel, for me anyway, was Ralph Swick, who made the pitch for using semantic web techniques to manage data in formats geared for interchange. At times it was hard to see the difference between semantic web and enterprise 3.0. I don't entirely agree. I just can't see enterprises throwing much energy (and money) behind the semantic web until the ROI picture becomes more clear. Perhaps its gaining ground on the web, where there's more data, but not behind the firewall.
My position was that Enterprise 3.0 will be dominated by context and insight. Many of the E20 tools we use now are feeding data silos. I gave an example of an E20 pilot we are doing with the internal software development community. The platform wasn't even really up before developers were asking to integrate the conversations into their development tools, or the other way around. Microblogs and wikis are great tools, but they are endpoints.
The point I am still noodling on is my assertion that search will be the primary driver of Enterprise 3.0. We're starting to make progress with a new search engine from Attivio. We're actively looking for opportunities to pull structured and unstructured data together to form insights (I hate the word "mashup"). The next wave will be the intersection of contextual data + transactions with search driven insights. I gave the example of sentiment analysis - today our marketing teams use tools to measure the mood of social interactions around different elements of our brand. Meanwhile, the sales teams are using standard CRM tools to manage their accounts. Sales is already asking to enrich their tools with new collective intelligence data; think salesforce.com Chatter. Enterprise 3.0 is when we start pushing sentiment data - our brand, customer's brand, competitive products - into their main workstreams. Using this example, sales would be able to interact around marketplace changes for our products or their customers.
Enterprise 3.0 is little more than a concept right now. Leave it to MIT to coin the phrase before the dust has even settled on Enterprise 2.0.
4. November 2009 20:23
I had a chance to spend a couple days at the SharePoint conference in Las Vegas. It was a surprisingly good trip and conference. Here's what I learned:
- SharePoint 2010 is going to be a very big release for that product.
- The best part is the business intelligence integration.
- The correct amount of time to spend in Las Vegas is 2 days.
The most exciting thing I saw was a project that was once called "Gemini", now called PowerPivot. PowerPivot is a plug-in to Excel 2010 and a server component that is part of SQL Server 2008 R2. I knew PowerPivot as an in-memory OLAP engine that supports very large datasets on the client. If you're working with large datasets, its is really damn impressive. What I saw at the conference was some of the back-end dashboards that support user and connection management. The dirty little secret of "self service BI" is that it can quickly turn into a "self service data warehouse", running on a wayward server under someone's desk. PowerPivot can't possibly solve the massive replication/export problem, but it at least makes it a little more manageable by adding easy to use metrics on the server. Administrators can view the amount of data and connections used by the various PowerPivot clients. That makes it easier to understand how much data is moving out of the managed base tables of a data warehouse into client managed "mashups".
here's a decent preview of the performance of the client:
One downside is that Microsoft is using its formidable powers of confusion and complexity to make the purchase process off-limits to all but the bravest licensing ninja's. One of the conference sessions was "licensing and SKUs". I realize how much it helps to understand the mass of jello that is Microsoft licensing, but I really wanted to go remind everyone that attended that session that there are less mind-numbing options in the other sessions.
more to come as PowerPivot matures and I get my CTP copy...
26. August 2009 21:44
a few days back I posted this on LinkedIn:
"Gregg Working on massively distributed problem solving, if anyone has any pointers..."
I clearly have some very smart contacts. I got some great responses. Its a bit tough to fit a decent question in 140 characters, so I thought I would explain a bit more...
In a nutshell, I am interested in the area between simple, repetitive tasks and discreet, complex problems (like open source software). I break it down this way:
- distributed computing architectures that support problem solving.
- problem decomposition using machine and human logic
- workflow systems that support parallel and hierarchical problem solving
- motivation in large, social systems
- characterization of participants (education, specialties)
there's probably way more to it, but that's pretty much what I am poking around at. There are some great examples in each...
26. August 2009 21:34
this is just a proof:
I don't think its the most interesting, but I think it fits the blog better. If I was planning a blog about wakeboarding at lunch the other would be great, but I'm just not that cool.
14. August 2009 00:02
what's the first thing you decide when you buy a car - what color is it!
I need to decide a logo for this blog. I have three front runners, and I want to choose one this weekend. If you see one you like, just say so in the comments and I will go from there.
this blog is going to be about life and systems in large corporations - enterprise 2.0, distributed development, cloud computing and the like.
The logo will replace the neato little blue square at the top of the page.