A fault was simulated that would take an experienced engineer a couple of hours to troubleshoot and solve.
This AI solved it live in seconds from analyzing splunk logs and the configuration of devices. Applied the software fix autonomously (with the option to back out), and then generated the service report.
All the human did was to ask it to handle a servicenow case id.
—
The future is not general AI like chatgpt, but smaller models that are highly specialized and highly capable in that aspect.
Think of these models as the equivalent of reusable subroutines in programing - except vastly more flexible and vastly more capable. Instead of passing args, their training provides the return based on their evaluation of your unique needs.
Already being done. From 1 dozen human agents here to 3. These 3 agents in the future won’t need to talk to server, client or network experts. The AI will replace those experts.
That's considered cheating. Revealed a lack of ethics and integrity. Passing AI work as his own. People with such a character will continue to be dishonest. It's best to sack him.
That's considered cheating. Revealed a lack of ethics and integrity. Passing AI work as his own. People with such a character will continue to be dishonest. It's best to sack him.
It isn’t actually. It’s working efficiently as the Human in the loop aka QC. This is the correct way forward. Leveraging AI to vastly improve throughput.
Now if the fellow passes bad/unvetted code, they’ve failed and need to be written up.
It isn’t actually. It’s working efficiently as the Human in the loop aka QC. This is the correct way forward. Leveraging AI to vastly improve throughput.
Now if the fellow passes bad/unvetted code, they’ve failed and need to be written up.
This forum is moderated by volunteer moderators who will react only to members' feedback on posts. Moderators are not employees or representatives of HWZ. Forum members and moderators are responsible for their own posts.
