6 ms·

As a felon myself working in tech, I find this attitude refreshing. It could work in IT...but would require significant changes to how things currently work. I

As a felon myself working in tech, I find this attitude refreshing. It could work in IT...but would require significant changes to how things currently work. I can't say for sure whether it would be worth it to restructure entire organizations to benefit the disenfranchised, but my instincts are telling me that it would add a reddish hue to the bottom line and thus is unlikely to be adopted. I do think that this sort of attitude could add value and blacken the bottom line if applied to many other types of industries (eg. construction, manufacturing, machining, food service and much more).

Kalium 8y ago
I suspect this practice would take the expected time from hire to productive from maybe 1-3 months to 1-3 years. This seems likely to be an expensive prospect, and that's if we assume every single person makes a good engineer. Based on the comments of others, I suspect this approach is most useful in a business where you have positions at a variety of skill levels. In IT and tech, the base skillset required can be substantial.
- ThrustVectoring 8y ago
  >if we assume every single person makes a good engineer. Not every person makes a good engineer. The US Army, with very good reason to train anyone who is willing, has decided that anyone with an IQ under 83 is not able to be trained to any position.
pc86 8y ago
If the overall effect is cost-negative, what is the incentive for a company to spend more resources to hire "the disenfranchised" (felons)?
- ilovetux 8y ago
  There would indeed be little incentive for a lot of businesses (specifically IT but also others). There are, however, other industries where this sort of career path already exists and with less rework could implement an "open-hire" program. Edit: I didn't really address your question. While there would be little incentive, there would not be none. For instance a better public image (like being an equal opportunity employer, but even more so) I believe they would end up with a more loyal workforce (but this is debatable and even if true it would be difficult to quantify) but... They might also upset existing employees who were hired under a more traditional process There would be less incentive to pursue higher education if this were common practice In short...I am tempted to look for a way to make it work, but also torn because I know that the needs of the business take a backseat to philosophical ideals. I do, however, believe that this could be implemented in certain industries and be a net-benefit to all
munchbunny 8y ago
I suspect with IT you run into a general trust issue, since IT often operates in an information-privileged environment. I would love to see better tools and technical infrastructure for requiring authorization to touch sensitive data, so that people in companies can operate with access to "the information you need but no more." Something like that would both protect customer privacy and de-risk hiring at the entry levels. It would most definitely make getting access to data tougher for lots of people, but as engineers who are often handling customer data, I think it's a generally more responsible (ethics-wise) approach.
- ilovetux 8y ago
  > so that people in companies can operate with access to "the information you need but no more." I think SELinux addresses this problem pretty well, but there is no real requirement to give access to developers for a production database when a development database could be setup with test data. I'm not arguing that this is a great idea for IT, but trust has been an issue from the beginning and there are more than a couple of solutions with varying degrees of quality.
  - munchbunny 8y ago
    > there are more than a couple of solutions with varying degrees of quality I think you raise a good point. If I were to phrase my wish better, I think it comes down to your implication that "varying degrees of quality" means some solutions do the job better (like SELinux) and some that do it worse. The more work it takes, the more you end up with startups where everyone has the passwords to the production database because it's just easier operationally. But when you do that, then you start to get into institutionalized risks like Uber employees stalking exes. I'm well aware that this is a case of engineering a solution to cover the 0.01% case, but when it comes to sensitive data, you're usually guarding against a tiny number of bad actors who could do a lot of damage. As an example, let's say you want certain developers to have access to non-sensitive parts of a Postgres database, and you want to be able to temporarily authorize developers with worthy exceptions to touch sensitive data, and you want to leave an audit trail. Let's say you're doing it this way because of HIPAA. It's currently a pain to do, and you'll have to write a bunch of automation to make it work. Now let's say you also have some S3 stores or ETL pipelines using Kinesis, and you have potentially sensitive data in there too. In the ideal case, you could manage the roles and groups and elevations in a more central place, federate those identities over to the Postgres cluster, SSH permissions, etc. That way data and privacy controls could actually be lightweight, which I think would be an overall win for the industry.

Kalium 8y ago

I suspect this practice would take the expected time from hire to productive from maybe 1-3 months to 1-3 years. This seems likely to be an expensive prospect, and that's if we assume every single person makes a good engineer. Based on the comments of others, I suspect this approach is most useful in a business where you have positions at a variety of skill levels. In IT and tech, the base skillset required can be substantial.

ThrustVectoring 8y ago
>if we assume every single person makes a good engineer. Not every person makes a good engineer. The US Army, with very good reason to train anyone who is willing, has decided that anyone with an IQ under 83 is not able to be trained to any position.

pc86 8y ago

If the overall effect is cost-negative, what is the incentive for a company to spend more resources to hire "the disenfranchised" (felons)?

ilovetux 8y ago
There would indeed be little incentive for a lot of businesses (specifically IT but also others). There are, however, other industries where this sort of career path already exists and with less rework could implement an "open-hire" program. Edit: I didn't really address your question. While there would be little incentive, there would not be none. For instance a better public image (like being an equal opportunity employer, but even more so) I believe they would end up with a more loyal workforce (but this is debatable and even if true it would be difficult to quantify) but... They might also upset existing employees who were hired under a more traditional process There would be less incentive to pursue higher education if this were common practice In short...I am tempted to look for a way to make it work, but also torn because I know that the needs of the business take a backseat to philosophical ideals. I do, however, believe that this could be implemented in certain industries and be a net-benefit to all

munchbunny 8y ago

I suspect with IT you run into a general trust issue, since IT often operates in an information-privileged environment. I would love to see better tools and technical infrastructure for requiring authorization to touch sensitive data, so that people in companies can operate with access to "the information you need but no more." Something like that would both protect customer privacy and de-risk hiring at the entry levels. It would most definitely make getting access to data tougher for lots of people, but as engineers who are often handling customer data, I think it's a generally more responsible (ethics-wise) approach.

ilovetux 8y ago
> so that people in companies can operate with access to "the information you need but no more." I think SELinux addresses this problem pretty well, but there is no real requirement to give access to developers for a production database when a development database could be setup with test data. I'm not arguing that this is a great idea for IT, but trust has been an issue from the beginning and there are more than a couple of solutions with varying degrees of quality.
- munchbunny 8y ago
  > there are more than a couple of solutions with varying degrees of quality I think you raise a good point. If I were to phrase my wish better, I think it comes down to your implication that "varying degrees of quality" means some solutions do the job better (like SELinux) and some that do it worse. The more work it takes, the more you end up with startups where everyone has the passwords to the production database because it's just easier operationally. But when you do that, then you start to get into institutionalized risks like Uber employees stalking exes. I'm well aware that this is a case of engineering a solution to cover the 0.01% case, but when it comes to sensitive data, you're usually guarding against a tiny number of bad actors who could do a lot of damage. As an example, let's say you want certain developers to have access to non-sensitive parts of a Postgres database, and you want to be able to temporarily authorize developers with worthy exceptions to touch sensitive data, and you want to leave an audit trail. Let's say you're doing it this way because of HIPAA. It's currently a pain to do, and you'll have to write a bunch of automation to make it work. Now let's say you also have some S3 stores or ETL pipelines using Kinesis, and you have potentially sensitive data in there too. In the ideal case, you could manage the roles and groups and elevations in a more central place, federate those identities over to the Postgres cluster, SSH permissions, etc. That way data and privacy controls could actually be lightweight, which I think would be an overall win for the industry.