Previous month:
March 2004
Next month:
May 2004

Do Conversations make any difference?

Dave Pollard, How to Save the World

This followed on from Dave Pollard's thinking on Presentations.

Now I've read it, I won't dare say anything for days.


Minutes, anyway.

Lately most of my meetings have been one-on-one, so I've started to look at conversations with the same skeptical eye as presentations. How much do we get out of them really, and are they truly about communicating or actually about something else? So far I'm just listening to others' conversations, whenever I get the opportunity. Since I'm male, you will appreciate that this is very difficult for me to do! But I'm also finding out (as most women already know well) that it can be very entertaining, if you pay attention to the whole conversation and not just to the words being said.

I'm starting to think conversations are as useless a medium for effective intellectual communication as presentations. It's too early for me to present any unscientific conclusions, but here's what I've observed so far -- I'd love to hear what you think about all this:

Linguistics professor Deborah Tannenseems to have a valid point when she says women and men (with some notable exceptions) converse in entirely different ways, and they converse differently with members of the opposite sex than with members of their own.

Conversations have a myriad of complex but unspoken cultural norms, styles and rituals (taking turns, pausing, nodding, apologizing for interrupting or misunderstanding etc.) When two people with different norms, styles, or rituals try to converse, or when a third person ignorant of the styles or rituals shared by the other two tries to enter a conversation, the result is both comical and tragic. A form of violence, even.

Most people don't appear to listen to what they themselves are saying. Many conversations include someone saying "I didn't say that" when in fact they did. I suspect if people listened to a tape or video recording of their conversations they would be stunned. They might never say anything again!

Most of the real communication in a conversation is not in the words. It's in the nuances of body and eye language. It's in the tone of voice. It's in the pauses. It's in the physical proximity or distance of the conversants.

Many effective conversations appear to be really interviews. That entails specific roles for the two conversants, with the interviewer's role being the more difficult and more important. If one person is mostly asking questions and the other person is doing most of the talking, it's an interview, not a conversation.

Conversations with more than two people are generally either parallel sequences of two-person conversations, or moderated conversations, where one person is clearly directing the conversational 'traffic'.

Conversations would, I think, be much more effective if we had a ritual of having each conversant state upfront what their personal objective for the conversation is. I appreciate that in some cases this must be done tactfully: "I've wanted to meet you since Mr. A told me that you... ", or "I'm looking for some help with..." In the absence of such a protocol, a lot of initial conversations exhaust an enormous amount of participants' energy trying to figure this out tacitly.

From watching online chat (the only written medium that in my opinion is fast and immediate enough to really qualify as 'conversation') and listening to young people especially talk, what people seem to want most from conversation with friends is reassurance. Everyone is always fishing for compliments and confirmation, and, unless and until they clearly know and trust the offerer very well, dubious of the offerer's motivation when they get them. Few people, it seems, are really looking for advice, debate, or 'constructive criticism' in a conversation. But many seem enthusiastic to offer these things anyway!

You can tell almost immediately whether participants in a conversation trust each other or not. If you want to observe conversations where there is trust, go out for dinner a lot, and avoid offices and bars.

Follow me on Twitter: @IanYorston

Do Presentations make any difference?

Dave Pollard. How to Save the World

This struck a chord.

My last important learning before I left Ernst & Young was the astonishing discovery that almost none of what business presenters say gets 'correctly' understood, internalized, or learned by their audience. By 'correctly' I mean what the audience thought the message was, is almost always radically different from what the presenter intended the message to be.

I base this conclusion on entirely anecdotal evidence: Throughout 2003, as a result of consternation about how so little of my presentations was sinking in, out of curiousity I began systematically debriefing with a few audience participants in each presentation I attended (whether or not I had been one of the presenters), as soon as possible after the presentations, and then fed back to the presenters what the audience said. The result was usually anger or stunned disbelief. Here are my totally unscientific findings from this 'research':

Regardless of the length of a presentation, audience members will recall no more than one important message or significant finding from a presentation, and, unless it is reinforced later, will forget even that one message or finding in about one week. They'll retain impressions about the speaker, but not what was said.

The only time a majority of the audience agrees on what the important messages or findings were, is when one or more of the following occurs:

  • the message/finding is emphasized at the very beginning and/or very end of the presentation
  • there is significant two-way conversation about the message/finding during the presentation
  • the message/finding is said repeatedly during the presentation, ideally by more than one person
  • the message/finding is conveyed by means of a story, joke, example or anecdote

In the absence of one of the above four 'aids', the probability that more than a small minority of the audience will understand what the presenter was actually trying to convey is close to zero.

Powerpoint slides with bullets, artwork & photos don't help understanding or retention. Charts, tables and 'top 10' lists can help, but only if they're simple, elegant, compelling, useful to keep, and properly explained.

Follow me on Twitter: @IanYorston

Shoppers shape up with the keep-fit trolley

Telegraph Connected

There are many ways a trip to the supermarket could be made more enjoyable. Combining it with a workout at the gym is arguably not one of them. Still, in a move that takes the notion of multi-tasking to a new dimension, a supermarket chain is now merging these two mundane chores of modern life in the hope of wooing time-pressed shoppers. Thanks to a new shopping trolley, customers of Tesco will soon be able to burn calories while buying the weekly groceries. Using technology normally found on gym machines, the trolley can be programmed to make it harder to push, increasing the heartbeat and exercising muscles in the legs, arms and stomach. A 40-minute shopping trip could burn off 280 calories...
Follow me on Twitter: @IanYorston

Use your LOAF


LOAF is a simple extension to email that lets you append your entire address book to outgoing mail message without compromising your privacy. Correspondents can use this information to prioritize their mail, and learn more about their social networks.


LOAF creates and maintains a database of all your correspondents, defined as people to whom you have sent email at least once. Every time you send an email message, LOAF appends this information to the email message, using a format described further below. LOAF-enabled correspondents collect and store this information in their own local databases.

When you receive an email from an address you have not previously written to, LOAF checks to see if the email address is known to any of your existing correspondents. This essentially sorts incoming email into three categories:

1. Mail from complete strangers

These are people whom you do not know, and who are also unknown to your correspondents.

2. Mail from partial strangers

These are people you have never sent email to, but who have gotten email from at least one of your own correspondents. This email may deserve more attention, since at least one of your correspondents took the time to write back to the person.

3. Mail from people you know.

This last category consists of people whom you have written to before. Presumably this is email you're most interested in, unless it's another forward from your mom.

Mail in category (2) can be further classified by counting how many correspondents you and the sender have in common. If the originating email appears in the address books of several of your correspondents, this may indicate a person with whom you have many connections. Insert standard social network theory here.

Follow me on Twitter: @IanYorston

Security Quotables

Don Park

More from Don Park

These are just catchy sentences floating in my head so I thought I should dump them here:

Best way to remove a threat is to make it worse.

Best way to protect a secret is to not have it.

If you don't know it, you don't have it.

Protection most appreciated is visible protection.

Invisible protection makes users more gullible.

Follow me on Twitter: @IanYorston

Zombies at Starbucks

Don Park

Another excellent piece of thinking from Don Park.

This particularly ghoulish scene from the movie Security Scenarios from Hell has three actors: WiFi, Zombies, and Spyware.

The perils of WiFi are well known and well publicized. WiFi is now commonly available which means those perils are now common as well.

Zombies are also well publicized. Typically, they are poorly protected servers or home PCs with broadbands which are hijacked by hackers, supposedly even traded like Yu-Ki-Oh cards in the hacker community, and used to increase scalability to their attacks and to reduce likelyness of capture.

Spyware is software running on desktops that monitors user activities and report back to it's master. Most of them are just privacy violators, some are used for more sinister purpose and are called trojans. Earthlink recently claimed that PCs had, on the average, 28 spyware installed. While I think the claim is over-hyped to fit their agenda, spyware is nonetheless common place and it's not difficult to place one on anyone's compure. If your PC is more than six months old, chances are that there were plenty of opportunities for hackers to seed it with spyware.

So here is the scene: imagine a new class of spyware that monitors wireless network packets using code from these open source wiretapping tools. AirSnort and one of the ARP poisoning packages should be enough. Now imagine this spyware being delivered to laptops with WiFi cards that supports features AirSnort needs. The laptop just became a new kind of zombie, which I call wireless zombie, that only wakes up when the WiFi card is used.

All that is missing from the scene is the stage: a WiFi hotspot like Starbucks. The laptop owner sits in a corner and access the Net through the WiFi, it could even be someone like me writing this very blog post. The spyware wakes up and starts monitoring the wireless traffic looking for passwords and credit card numbers. If very strong encryption is used, wireless zombies can form a global grid and split up the work of cracking encryption keys. Once a month, the zombies reports back to their master via USENET posts.

This Zombies at Starbucks scenario is particularly nasty because the potential number of compromises is just staggering. Maybe the FCC will have to dictate higher level of standards and send out a warning that helps WiFi users detect wireless zombies by the unusual fan activities triggered by the zombie grid working overtime.

Follow me on Twitter: @IanYorston

File Formats on Google


Useful info... but particularly note the closing observation.

Andy notes that Google is now indexing Flash files. Search for "skip intro" to try it out. Upon seeing this, the gray-bearded conspiracy theorist in me wondered if Google was unfairly promoting the Flash format over Adobe's competing SVG format in order to crush Adobe into dust. I needn't have can search Google for SVG files just fine (because they're text files).

Of course, you can search Google for all kinds of filetypes, text and otherwise: .rdf (RSS, FOAF, etc.), .xml (RSS, Atom, etc.), .torrent (BitTorrent), .aspx (.NET), .php (PHP), .csv (comma-delimited data file), .vcf (vCard...look, global address book made easy!), etc.

Follow me on Twitter: @IanYorston

New Spin On Computing

Information Week, Antone Gonsalves, TechWeb News, April 26, 2004

IBM and Stanford University on Monday launched a joint research center focused on an area of nanotechnology called "spintronics," which could one day end the irritating delay people experience when they turn on their computers.

Spintronics involves use of the spin property of electrons--tiny particles in atoms that produce electricity when flowing through a conduit. Controlling the spin of electrons within a computer's CPU, the chip that provides processing power, is how researchers hope to create the fast-loading computer as well as other enhancements.

IBM and Stanford disclosed the formation of the Spintronic Science and Applications Center, or SpinAps, at IBM's Almaden Research Center in San Jose, Calif. Spintronics research will be conducted at Almaden and at Stanford's labs. However, IBM and Stanford researchers don't expect to see commercial products using their work for five to 10 years.

Most of today's electronic research focuses on the ability of electrons to carry an electrical charge. By focusing on an electron's spin properties, IBM and Stanford researchers hope to make breakthroughs in chip design, an IBM spokesman said.

The new design involves stacking layers of material, two or three atoms thick, to control the spin of electrons as they travel through the layers. Among the benefits would be the creation of magnetic random access memory.

RAM is where the computer loads the software needed to run when the machine is turned on. It's this process that takes time. In addition, when the computer is turned off, everything in RAM disappears, which means the process has to be repeated.

Magnetic RAM, on the other hand, would remain on, even when the computer is shut off, which means computer-launching software would be saved. As a result, turning on a computer would immediately take it to its previous state.

"RAM is volatile, which means if you shut down the power, then the information is gone," the spokesman said. "What spintronics allows you to do is set state through electron-spin interaction, so (the chip) doesn't need power to keep the information."

In addition, magnetic RAM won't leak power like the RAM used in today's computers. Because of power leakage, computers have to constantly reload RAM with the software needed to keep the machine running.

"It's like having a leaky bucket, and you have a faucet on to keep the level of water the same," Ross said. "The power is like the faucet. You have to keep it on as the power leaks out."

Eliminating this inefficient power usage means a laptop will run longer on its battery, and computer makers would be able to find new ways to use the power that will no longer be needed for RAM, Ross said.

Spintronics is a field within nanotechnology, the science of developing materials at the atomic and molecular level in order to imbue them with special electrical and chemical properties. Nanotechnology is expected to make major contributions to the fields of computer storage, semiconductors, biotechnology, manufacturing and energy.

Follow me on Twitter: @IanYorston

Soldering iron added to nano-toolkit

New Scientist

An atomic-scale conveyor belt may also be the smallest soldering iron ever created. The new device, which ferries molten metal, is made from carbon nanotubes just 20 millionths of a millimetre in diameter.

The discovery could pave the way for nano-machines that are pieced together from smaller components, rather than emerging from chemical reactions.

"There has been a dream for many years to build nano-structures piece by piece, like building a large-scale machine," explains Alex Zettl, who built the nano-soldering iron at the University of California in Berkeley. Then the structure is no longer constrained by the chemistry of its components, he says.

Currently, nano-probes can nudge atoms one at a time from one place to another. But to generate the flow of molten material necessary to solder parts together, hundreds of thousands of atoms must be moved.

Zettl begins by spraying pure carbon nanotubes with gaseous indium. The metal then condenses into solid droplets between one and 10 nanometres wide. Using a nano-manipulator built in his lab he connects the tubes to a circuit and applies a small voltage.

Heat from the resulting current melts the droplets, which scurry along the nanotube's surface and collect as a bubbling liquid at the negative end. Reversing the voltage shunts indium to the other end, meaning the movement is driven by electricity rather than a thermal effect. And varying the voltage changes the flow of liquid indium from a "drip-drip" to a "surge", Zettl says.

The liquid bubbles could be wicked off the ends of the tubes and used to solder tiny parts together, he says, though this has yet to be demonstrated.

Zettl is unsure why neutral indium atoms move in response to a voltage. "It looks like the indium is being positively charged," he says. He suggests that electrons may be migrating from the indium to the nanotubes.

Droplets of gold, tin and platinum can also be made to move from one end of the tube to the other. "It's the first time we have controllable, atomic scale motion of mass," Zettl told New Scientist.

Follow me on Twitter: @IanYorston