I've been using GitHub Copilot for some time, and while it's decent, my company is really pushing for AI integration, and we all have licenses for various tools. I appreciate the potential for using AI to brainstorm ideas and the cool agent mode that helps navigate files to answer questions. However, I'm finding it hard to get on board with the industry's shift towards agentic implementations. I'm in my mid-30s, so I'm not trying to sound like the old-timer saying 'get off my lawn'; rather, I want to create something useful that I won't regret in six months. My concerns include: 1) Watching a video about using AI for customer service, I worry about the tendency for these models to make mistakes or 'hallucinate' information. I'm not comfortable with relying on a black box. 2) I've read about model poisoning, which is often linked to political motives and could cause significant issues over time if an AI's underlying data changes. 3) I just don't get why we're pursuing AI solutions when they often seem more costly than simpler alternatives like a basic web form. Am I wrong to think the technology is cool, but the approach may not be the best? Companies like Shopify are pushing to eliminate jobs unless they can prove they can't be done by AI, and I want to prepare for this new reality without losing sight of what makes sense and what doesn't!
4 Answers
While there are undeniable security risks associated with AI, these are typically the domain of security teams. The reality is that AI is about investment; the technology is evolving and will become more efficient and cheaper to run over time. Practically, it can be a useful ally that’ll really help organizations scale—if they do it right.
True! AI is an investment in our skills and capabilities and will improve over time. The key is to strike a balance and get the foundational aspects right.
The ongoing support and maintenance of AI systems is a hidden cost that many underestimate. If you’re in mundane roles, you’re replaceable now – AI is just the latest tool companies are latching onto to cut costs. Some roles will always need a human touch, but it’s definitely creating a wave of concern among entry-level positions.
You’re spot on. If companies aren't careful, they risk creating a future workforce with fewer experienced individuals to handle actual issues when they arise.
A lot of organizations are making a big mistake by thinking 'everything must be AI.' The focus should be on automating mundane tasks that nobody enjoys. AI could really excel in areas like data entry or processing receipts, where natural language is challenging but useful. It’s all about solving actual problems, and if used in the right context, AI could create a lot of efficiency without overcomplicating things.
Exactly! AI shouldn't replace every role; it should assist where it makes sense. There are many boring jobs out there that people want to escape – that’s where the real opportunity for AI lies!
Honestly, a lot of this AI hype is just fueled by money. Leaders think they can slap 'AI' on a project and it’ll solve everything, but often it won't deliver. Sure, AI is cool and all, but it feels like people are getting blinded by the potential profit instead of focusing on real use cases that might actually work well with AI. It's like a trendy bandwagon roll that doesn’t necessarily mean you make good music!
Couldn't agree more. It's like everyone wants to get on the AI train just to look good, but whether it genuinely helps is another story. Some companies are just seeking quick bucks, and most techies can spot the smoke and mirrors.
Right! It’s hard to watch all this overblown hype without thinking about the genuine utility of LLMs. Some aspects are quite helpful, but you have to sift through a lot of nonsense.
That’s all well and good, but continually relying on AI-solving problems could lead to disastrous outcomes if not approached cautiously. There's a fine line between leveraging tech and losing control.