IBM and Numerati Revisited

Recently I posted my thoughts on 'IBM and Numerati'. I got several interesting comments on that post. And I was pointed to posts by Luis Suarez (mind you, an IBM-er) and Dave Snowden (an ex-IBM-er). I had already read them before posting my ideas on this topic, but didn't respond to their posts.

But too be clear. I was just as surprised to read about this topic as Snowden. And funnily, an IBM-er like Luis says this is not common practice at IBM either...

However, as I said, I do find the idea intriguing... and scary. True, like Luis and Ton Zijlstra commented, this is different from web 2.0. For me too, I like the web 2.0 approach more than the Numerati approach.

But I do find this approach closer to web 2.0 than my commenters. Luis says:

IBM is not routinely analyzing employees e-mail, calendars and chats without employees' permission or knowledge

But how does this relate to IBM Whisper and the opt-in 'Social Network tool' IBM runs internally. I posted on these two here. I wrote:

Erik goes on to show IBM Whisper: it is an automatically generated list of people you know and/or subjects you are interested in. It comes up with suggestions for documents, links, articles, etc that you might find interesting. (...)

Erik shows the Social Network Analysis tool IBM has in which email communication and Instant Messaging is tracked. Employees can choose to opt in or not. Type in a word you want to see the network for. The tool comes up with a diagram and/or map with connections between people on it.

Yep, these are opt-in, so employees have to give permission. But, as far as I understand the Numerati (again, I haven't read the book yet), it's not so secretive either. I mean: they're writing about it in Business Week!

So, if it's not secretive and employees can agree/disagree to opt-in, where does this bring us? Could this approach help us like the Techmeme is helping us find interesting stuff and be more productive? Or, in other words, is IBM Whisper and their Social Network Analysis tool helping the employees that use it be more productive? I agree with Ton, that you would always want to tell 'the machine' what to process and what not. It should always be your choice. But even then, relating to an older but interesting Wired article about radical transparency: if we would all be radically transparent (in the organization), would this help us make the company more productive, etc. (And there are limitation to transparency. We can also make things transparent that we can write down.)I mean, if you ask older people, they think your crazy to write out in the open (blog), tell all what you're doing (twitter), etc. And we look at them in a funny way and say: We're more productive than you are, just watch us and join us! Could the Numerati or (even more) transparency be the next step?

I'd love to hear what you think!



  1. Hi Samuel! Great follow up! Thanks for putting together these thoughts and for sharing your additional input! A couple of things to add further up into the conversation...

    1. The quote you mentioned above is an IBM statement, not me. It was IBM's positioning towards the whole topic of that particular piece of research. However, I must say that I wholeheartedly agree with it and go by it as well. And it makes me feel pretty good about the fact that my privacy is respected at all times and in all circumstances.

    2. With regards to your comments on Whisper, I think the keyword here is "opt-in" model. Basically, people voluntarily sign up to allow Atlas / SmallBlue to crawl that IM & email data, but without the consent from the individual it would not happen.

    Inside IBM I can tell you how plenty of folks have given their consent and have jumped into it, more than anything else, because it allows them to reach out to a wider number of experts, but from that to say that every single individual's email and IM conversations will be tracked that's just a bit too far, specially if not following the opt-in model.

    Like I said, Whisper has got tremendous value in how it helps find experts a lot easier, but it is also true that there is still a limit on how private & individual information gets treated and certainly knowing that it would always be "me" in control of that flow surely makes a difference.

    Sometimes you have to take a stand between privacy & security and how information is gathered / captured and I am glad IBM, as one of its employees (Insert usual disclaimer here please), treasures my own privacy just as much as I would want it for myself.

    Thanks again for the thoughts, Samuel!

  2. Hi Samuel,
    Interesting conversation you have here. I just wanted to add my two cents. I'm the author of the Numerati and of the BusinessWeek excerpt of it. IBM was wonderfully cooperative throughout the reporting. They gave me great access to researchers and I did many hours of interviews. For that I'm grateful. Following the publication of the excerpt, IBM asked me to make clear in my talks and writings about the book that the company collects and analyzes workers' data on an opt-in basis.
    Moving away from IBM, I must say that companies will increasingly be analyzing their workers' data, including e-mails, and many are already doing it. For regulatory and security reasons, companies have a responsibility to sift through communications looking for leaks of insider information, as well as pornography, sexual harassment, and of course, spam. So in most companies, our mail is being automatically analyzed. The next question is whether this process should be widened to analyze the social networks within the company. This, at least in theory, would help create better and more efficient teams. The next step would be to analyze subjects, word patterns, syntax, etc. Yes, there are privacy issues here. Beyond email, there are plenty of tools to analyze how workers are making use of software in their computers. The information about worker behavior is cascading. And the pressure to carry out sophisticated analysis will grow, I suspect, during this economic downturn. I, for one, assume that my email and online behavior are being monitored. Best regards, Stephen Baker

  3. Hello Steve, Great to hear from you and wonderful that you're listening! Thanks for the clarification. Looking around at other blog posts on your book and the comments on my post, I think I'm more open to the road you've described than others are. I'm really curious where this will take us. It doesn't all have to be bad, is what I'm thinking. Although it sounds scary, it could have big benefits too. I'm following your blog and look forward to new insights in this area.

  4. Snowden paraphrases - Imagine that your boss finally recognizes your strengths,then he puts you into situations where you will thrive.

    In enterprise 2.0 I imagine it becomes an autonomous ecosystem where after a while, information and people I have interest in, always comes to me (due to networking with like people), we sensemake or filter the enterprise through our network of friends...not only do we subscribe to each other and converse, but we altruistically send each other information that we may be interested in as we have a trust and reciprocating relationship.

    We all have skills and interests and when we connect with people we get lots more opportunities coming to us that suit our interest and skills. Imagine in your employment you were free to go from job to job within a company, kind of responsible for'd would be best placing yourself in where you would thrive...and you find these great jobs because you are connected.

    This is quite utopian because not everyone may want to be involved in a social network and carve out their own opportunities...some people like email, they just want to go to work, put in 8 hours and go home...enterprise 2.0 is not for everyone, you shouldn't have to take part if you don't want to.

    Another reason this is utopian is that currently we have to live in a practice or unit, often we get seconded, but still we have a home. In a utopian enterprise 2.0 all employees would be like inhouse consultant/contractors, making their own work...but what happens in idle periods, you still need to be able to charge to an business unit overhead cost code.

    Anyway, Luis makes clear this meme is not about privacy as you have to opt-in.

    What I want to say is why does my boss have to place me into situations where I will thrive, when I can ask him myself. If my boss knows about some tasks that would suit me that I don't know about, well then my boss can network with me to let me know, just like anyone else in my network lets me know about tasks that I will like to get involved in. I still think I will create my own opportunities and I have my own social network.

    Ton is right it's like Taylorism for the knowledge age.

    "...once the workers are represented as mathematical models, it will be far easier to break down their days into billable minutes and send their smarts to fulfill jobs all over the world."

    This is totally top-down, what if I don't agree with a suggestion that had been computed for me...what then, I will be defying many times am I allowed to refuse before I get fired.

    You see where this is going, it's an all or nothing thing, there is a directive, and that's efficiency, and if you are a spanner in the works, you are going against protocol.

    Just like the cashless society we are moving into, everything is moving to global currency, id cards, and soon enough RFID chip impants...people are becoming commodified as consumers, we are losing our humanity. If you are a rogue your chip gets turned off, and you can't survive or function.

    I think this relates to this discussion as we are talking about a top-down model where you are being monitored and profiled.
    Humans don't like the idea of this, they like a flat field where we all network...

    I understand the idea is to put you into positions that you will like, so it's seen as top-down friendly, but then Amazon doesn't always recommend me books I like...knowing someone mathematically is just not the Driessen says not all a person's know-how is digitally documented

    The problem with Atlas/Whisper is that it seems to monitor place like email and IM that are both private/public...there is no way I would do this. Why not monitor my public contributions like my blog, wiki, bookmarks, micro-blogs, contact list, etc...

    If management are looking for the right person they can browse the expert tag cloud, they can search the blogs and networks, and go from there.
    Now both management and workers can find experts other than word of mouth.

    But to turn this into some sort of perpetual mining/monitoring SNA gone mad, is not comfortable.

    The instant rebellion is to stop blogging so I can't be profiled, and in turn, if all did this, this breaks down the system as it's based on there has to be content to be monitored.

  5. Sorry forgot to add this in, but in the end you wouldn't even need a boss.

    A comment on Snowden's post says:
    "...what would be the added value of the boss? If the best actions are optimized "mathematically", then the boss only makes them worse. Plus their salary is a waste of money"

    There would be some machine that would tell us what to do based on our past behaviour of our digital selves. When I say digital selves, I mean the machine doesn't really know us, it only knows a portion of us based on our digital content...and even this it doesn't "know", it attempts to piece together a coherent picture of a person...but I don't like to work in an apple factory, I want to work for apple computers...this is a parody, but you can see there would be lots of other semantic issues.

  6. Well, it can sound terribly top-down, and in some cases I'm sure it will be. But if you look at the information as suggestions, and not mandates, it might be OK. One point for IBM and other big companies is that there are thousands of people working in lots of countries, inside and outside the company, and it makes sense to index them by all different metrics and make them findable.
    Stephen Baker

  7. Interesting discussion! I understand both approaches. I was studying semantic web/web 3.0 yesterday. The semantic web can only live on smart things people have shared on blogs, wiki's, the internet. But the semantic web promises to automatically connect the bits and offer you possible new insights. This could be scary, but it could be very helpful too. And I don't think the semantic web will be opt-in...

  8. as Read more about Green eco architectural designs are reusable materials, green designs etc.. ..… Create an Eco friendly Green design… Save Earth.." as of natural materials with almost modern concepts using A few quick ways to reduce your personal carbon footprint: employ eco-friendly methods of travel, eat, minimize use of plastic products and save electricity, water, paper and use of non-toxic alternatives Carbon Footprint is a measure of the impact human activities have on the environment in terms of the amount of greenhouse gases save earth use architects suggested materials


Post a Comment

Please leave a comment! Just log in using one of the formats and if you want me to get back to you. Otherwise comment anonymously.

Popular posts

Keep the Intranet Small

Enterprise 2.0 Research

Innovation in Turbulent Times