So not even the Microsoft execs are using this shit.
View: https://x.com/WindowsCentral/status/2001631280077963360?s=20
My company (in Healthcare IT and other vertical markets) is promoting "AI: Every Role, Every Day" and has a regular webinar series where people highlight what they're doing with whatever tools fit their role (all Devs have GitHub CoPilot), demo, Q&A, etc... big push from mgmt, and a bunch of documented successes so far but a lot still TBD.
We get a weekly memo with info on the projects that are getting attention at that level... Big variety but often specific processes (tools or generating documentation, marketing, or other meta-products) where time saved can be measured against already known averages, or on my end some dev tasks like code maintenance analysis, initial code review to catch the big stuff and reduce time that seniors spend on reviews, validate against standards, etc. Still a long time, and more than one leap in capability, to come before it would replace anything but a percentage of your junior devs, but I'm sure it's affecting hiring projections. To me right now it's more like just another tool we can use to be more productive like every other technological advance has been, though with more noise about it, for sure. I mean, I used to code in Assembler and debug with a kernel debugger terminal and a serial cable; people were scared then that the move to higher-level languages and design-level tooling would mean they wouldn't need programmers anymore. I'm more worried that these AI Tech Bros are going to crash the market and I have to wait longer to retire than I plan to, than I am that AI can replace the value I bring.do the poor schmucks realize they're helping to make their own jobs obsolete?
what are the successes, btw?
We get a weekly memo with info on the projects that are getting attention at that level... Big variety but often specific processes (tools or generating documentation, marketing, or other meta-products) where time saved can be measured against already known averages, or on my end some dev tasks like code maintenance analysis, initial code review to catch the big stuff and reduce time that seniors spend on reviews, validate against standards, etc. Still a long time, and more than one leap in capability, to come before it would replace anything but a percentage of your junior devs, but I'm sure it's affecting hiring projections. To me right now it's more like just another tool we can use to be more productive like every other technological advance has been, though with more noise about it, for sure. I mean, I used to code in Assembler and debug with a kernel debugger terminal and a serial cable; people were scared then that the move to higher-level languages and design-level tooling would mean they wouldn't need programmers anymore. I'm more worried that these AI Tech Bros are going to crash the market and I have to wait longer to retire than I plan to, than I am that AI can replace the value I bring.
Yup, definitely on the rise.. the tooling we have available within CoPilot is pretty cool and will help us crank out raw code faster.ok so yeah the usage for coding definitely seems to be real.
Yup, definitely on the rise.. the tooling we have available within CoPilot is pretty cool and will help us crank out raw code faster.
LOT of concerns... not the least of which is devs skill in prompting and giving the other levels of instruction to the AI tooling (the skill/engineering of it) to produce good output, not just output which "looks good". The tools themselves only know what they know, and generally only get you 80-ish% (to just pull a number out of my hat) of the way there.it makes sense that it would be very good at that but even then there's some concerns.
apparently windows 11 was done largely with AI and it's having all sorts of troubles, and that's just shortterm stuff and there could always be flaws hidden deeper in the code that end up breaking stuff later. And then if all the AIs start to rewrite code based on potentially flawed code....things could go south quick. But at the same time yeah this is one thing AI should be able to do very well.
Don’t get me started on new outlook. There is no way it was tested by humans, unless it was North Koreans trying to sabotage our productivity.it makes sense that it would be very good at that but even then there's some concerns.
apparently windows 11 was done largely with AI and it's having all sorts of troubles, and that's just shortterm stuff and there could always be flaws hidden deeper in the code that end up breaking stuff later. And then if all the AIs start to rewrite code based on potentially flawed code....things could go south quick. But at the same time yeah this is one thing AI should be able to do very well.
My exact experience too. It doesn't replace a brain. There is skill required to get the desired output, but if you know how to use it, the value that this tool brings is immeasurable. But yeah, I see it as something that can help increase productivity in employees rather than replace them. It's a tool. And it's an incredible game breaking one for that kind of thing.LOT of concerns... not the least of which is devs skill in prompting and giving the other levels of instruction to the AI tooling (the skill/engineering of it) to produce good output, not just output which "looks good". The tools themselves only know what they know, and generally only get you 80-ish% (to just pull a number out of my hat) of the way there.