Slashdot Mirror
Interviews: Ask Juan Gilbert About Human-Centered Computing
Awarded the first Presidential Endowed Chair at Clemson University, and being named a Fellow of the American Association for the Advancement of Science (AAAS), are just a couple of Juan Gilbert's more noteworthy honors. Juan is the Associate Chair of Research in the Computer & Information Science & Engineering Department at the University of Florida where he leads the Human Centered Computing Lab. With the help of students, the lab works on a variety of issues, including electronic voting, automotive user interfaces, advanced learning technologies, culturally relevant computing or ethnocomputing, and databases and data analytics. Dr. Gilbert has agreed to answer any questions you might have about computing and affecting society through accessible technologies. As usual, ask as many as you'd like, but please, one question per post.
Here are some questions for Professor Gilbert, regarding internet voting:
1. How will non-mathematicians know with certainty that votes have been properly received and counted?
2. If the security depends upon encryption, how will we know that encryption has not been broken by a secret agency with vast computing power? Further, how will we know that those involved in developing the encryption have not secretly offered back doors to such agencies, as has happened in the past?
3. What will a voter do if they experience an election-day denial of service attack?
4. How can we know that a vote has not been coerced if the voter votes from home (bullying spouse, etc...)?
5. What are the insurmountable difficulties with a paper-based election process that make internet voting desirable despite risks? Why is the United States no longer capable of counting cast ballots in public? It is clearly not the vast number of voters, since this is a distributed problem with a vast number of potential counters. What has become so broken among our pseudo-elites that this KISS approach is now considered so inappropriate?
To what extent are we able to compute safety related human dynamics issues and what is slowing us down in this particular programming area?
Can we ever come up with a safety system for a workplace that would be able to overcome employee buy-in issues early on, especially if the typical large corporation is in a constant tug of war with profit and employee needs?
You see whenever we introduce changes in policy in the workplace, employees assume they are going to be required to do MORE but they are not getting more money for the work so this tends at times to cause resistance from employees to safety policies. Management doesn't often understand the issues at hand so they tend to make contradictory safety policies as well, saying that things need to be addressed in a timely fashion.
But in the aftermath of this complexity, companies are often just faking safety in order to appear to be safe when in fact they are running at a significant moral hazard to everyone (their staff, the general public and anyone else for that matter).
This particular problem is of great interest to me and I find that whenever there is an imbalance between management and employee needs there is a systemic problem that is solvable but yet only once all the variables are on the table. The problem with human safety is that most of the variables are unknown.
The general equation for solving safety related issues is:
For every task an employee is required to do or will reasonably be presented with, the employee must be trained to perform the task safely within prescribed safety policy. This idea is fundamentally at odds with bravado in the workplace, hero complexes, profit margins and it goes directly against human psychopathy that is prevalent in modern corporate culture.
What's the best approach to stabilizing a safety model?
The dangers of knowledge trigger emotional distress in human beings.
Are computer interfaces getting finicky these days?
With mobile devices requiring a touch that must be
1. Precisely located when you can't see the target when touching it without transparent fingers,
2. Precisely timed in order to avoid being interpreted to mean exactly the opposite(long touch) of what is intended,
3. Perfectly still in all dimensions except the z axis to avoid being interpreted as moving the target instead of selecting it.
With mouse interfaces using "mouse-overs" to move, change, or otherwise obscure the original target.
With on-screen messages which go away in a huff if you don't react to them in some arbitrary amount of time.
Is the error rate of modern gui commands going up compared to older interfaces?
It seems to me that we want to simultaneously be able to prove to the voter that their vote was counted properly, while also wanting to ensure that the voter cannot prove to someone else that they voted a certain way (to prevent buying/coercing votes).
Adding to this...how do you ensure that the person voting is who they say they are, and not another family member, care provider, guest, etc.?
The Republicans love it because they can continue their mass voting fraud scams more easily, so any sensible person should be against Internet voting. I know in the neighborhood where I live, not a single person here admitted to voting for the war criminal McCain in the 2008 eleciton according to a door to door survey that Univ of Washington did, but McCain got 16% of the votes according the fraudlent mail-in voting. Here in WA we vote by mail, and there is massive fraud. There is no way 16% of the people in my neighborhood are that stupid. That proves it is a scam.
I misread the headline as "Ask Juan Gilbert About Human Centipede Computing".
How much impact will wearables have on the accessibility of technology to people with disabilities? A smart watch is a lot easier to move around with than a laptop. Do you think these smaller devices will have a big impact or is this just another fad?
I would guess most of us think that computers have existed at a time when much positive work on equality has been done. The internet is one place where it is fairly easy to be anonymous. Can you give an example of ethnocomputing concerns that aren't just economically or exposure to technology based?
From what I can tell most of the work done at the Human Centered Computing Lab seems to be from students. Since this means a fairly quick churn rate does this make it harder to complete big projects or does the constant influx of new blood make up for that?
I watched your TED talk about multiple teacher single student learning. Has any further progress been made on the project or have you integrated it into any real-life schools?
You've said that you created the Application Quest program in response to the Michigan affirmative action rulings to help colleges keep diversity without giving preference to race. Are any colleges using the system or expressed interest? Have you had an backlash over your stance that the issue really isn't about race but about capacity?
Most of the recent changes I've seen to driver controls seem wrong-headed. So many require the driver to look down at some screen or closely-spaced identically-feeling buttons. Only a few decades ago, car makers began moving functions to stalks to put them within easy reach, but the makers' usage is so different that it's more confusing than ever, particularly in this day when people are more likely to drive several different vehicles in a single day. I like steering wheel-mounted buttons, but now there are so many it has again become confusing, and again, makers refuse to adopt any standard placements/usage. Can we have programmable controls that follow a driver from car to car, always working the way that particular driver prefers? Must we resort to voice control? I despise talking to my car or any other device, including my phone. Is there no solution until cars can be controlled by thought?
A few years ago I read that some were working on systems where people's personal data is never decrypted yet some types of analytics can be performed on it. Has this work died or is it still ongoing and does it lead to anything useful from a privacy perspective?
With today programming languages, creating new new software requires learning a complex syntax with very specialized rules on how to combine words, even for creating very simple software (for example, web pages with trivial interactions such as folding and dragging items).
Some approaches to allow end users to build automated behavior exist, but they can only go so far. There are "drag and drop" interface builders for building web pages with forms, and graph languages for transforming data. But they only allow reusing pre-defined components which are built with traditional languages. Any behavior not supported by those components can not be added to the program.
There are also rule-based visual systems like Agentsheets that allow defining new behaviors without a strict complex syntax, but those are difficult to reason about when behaviors depend on several levels of nested rules.
My question is: what would be your preferred approach to achieve the goal of allowing end users build their own simple software programs? This assumes that we define "program" in a loose way, not necessarily in the traditional way but referring to any software artifacts for defining repeatable processes to handle information such as:
* building and classifying collections of related data, transforming the shape of parts of a document...
* or for automation of actions in time (turning on and off lights and engines at particular times or in a pre-defined pattern, sending messages to groups of people that follow certain criteria under some triggering condition)...
All this without requiring that the user learns a scripting language or otherwise needs to form a mental model of how exactly the program's execution evolves in time within the machine components.
Singularity: a belief in the "God" idea with the "demiurge" relation inverted.