We build a machine and ask it to make paperclips.
Technology has gotten cheaper, but other necessities haven’t in several decades, and likely won’t. Natural resources, for instance, are fixed or declining. [edit] Wages aren’t going to offset this either.
Are you guys curious about the Amazon store? I wonder if the customers are more or less social in the retail experience without the employees wandering around. I wonder how they react. Feels like a Black Mirror episode a bit.
+1 like you go girl
Allow me to interrupt this conversation by telling you that I appreciated your comment. Congratulations on having such fine wit.
Let us mark this occasion with the words taught to us by our host, ‘You go girl!’
Yeah, this got a good belly laugh out of me too. YGG!
Has she tried that again with her Kuka robot?
This does make me wonder how one armed people get peanut butter out.
That girl is hilarious but makes shitty robots on purpose. The robot arm is not the ideal solution for that problem, it’s more likely a big tube of peanut butter that gets squeezed out, a peanut butter roller, a hopper with bracing supports, etc.
Like the automated fast food restaurants of the future won’t be humanoid robots flipping burgers and dropping fries… The whole kitchen will be a custom built assembly line fed by giant hoppers.
What, like genetically-engineered crickets?
Yeah, and then they fry themselves for a delicious snack when they are nearing their inception date determined end.
It would probably be more attractive to genetically engineer cows to work as fry cooks and self terminate in the meat freezer.
No. Crickets are jumpers. Rabbits are hoppers.
Bah! Genetically engineered Hoppers FTW!
Do you guys think automation will eliminate IT jobs as well? I just don’t see how a machine could write code or diagnose and fix a hardware failure well enough to outstrip the demand for IT workers. Thoughts?
I would be shocked if we don’t get AI writing code soon enough. Though I’m not sure they would put the panache I put into their unit tests!
Amazon and Google’s server/admin ratio is bonkers - 1000’s:1, so it’s already happening to an extent.
It takes time as it is reliant on applications themselves being further obfuscated from the tin in design and deployment, such that any hardware failure does not matter as an app is designed with hardware failure in mind and to self heal through automated diagnosis and remediation. The automation involved in existing auto-scaling tools does part of this, app health monitoring tools do another part, continued use of virtualisation/containerisation to breakdown modern apps to a collection of microservices rather than fat stacks theoretically allows independent scaling and remediation of each service such that any individual failure is unlikely to take down a complete app and improving availability. I have little doubt AI is already being thrown at health monitoring tools to improve their ability to manage an environment with less human intervention.
But this will take time. ERP’s and CRM’s can have multi-decade lifecycles and massive stickiness. They are hard for big business to replace and have huge amounts of capital tied up in licensing, code customisations, business process integrations, such that the risks associated the business impact of a migration is too high, so status quo sits.
To some extent, they already have this capability. Watson can diagnose all kinds of crazy complex systems.
My understanding was that Watson is just branding now that the research project is done, for bog-standard rule systems or standard-framework ML.
Already happening. Our company is moving to “DevOps” (stupid damn buzzwords - I’ve been doing DevOps long before it adopted a new moniker), purposely so we can automate away the need for a lot of infrastructure folks. And for things that still need bodies (like coding) that’s being replaced by the cheapest outsourced labor we can find. The only people that will be left are ‘Business Analysts’ (ie., the english speaking folks that can act as the go-between between business and IT so business doesn’t know that their IT has been secretly replaced by Folger’s coffee. You know, except for the sudden increase in quality issues. But dammit, DevOps lets us fail fast! Quantity!).
A buddy of mine has worked in Cybersecurity for a few years (at FireEye, among other companies). They went through a big push to outsource their development a couple of years ago. He ended up doing a crapload of travel to India to supervise and get status. From my talks with him, it seems their experiences were mixed: the time difference made meetings and supervision difficult, and it seems the development teams in India work to the document with around 100% accuracy. So if your design document has functional gaps or has gaps - too bad, because you’re getting exactly what you asked for, and not a smidgen more.
As a result, they’ve dialed back on the outsourcing - they still use it, but it’s certainly not the panacea it was originally thought to be.
In my field (software development requiring a security clearance) there’s no chance of outsourcing, and so the rest of my career is pretty secure (not to brag, but in the DC area there is so much demand for cleared developers that I’ll get 5 to 10 emails a week for people trying to recruit me. I suspect it’s double demand if you have the “entry level” polygraph, and double it again if you have the max level poly).
@Daagar IMHO integrated DevOps is the new way of the world - and it’s awesome of course - it’s so much nicer to have control of our own build/deploy process as opposed to dealing with gatekeepers (a couple of years ago I worked on a big project where your changes could get deployed no faster than once an hour, and we had a dedicated Ops team whose goal was to deploy the webapps - that is, when they weren’t out at Wendys for hours at a time! If you were trying to nail down a bug it wasn’t unusual for it to take a day to get 3 or 4 good runs of the code so you could try to figure out what was going on).