I'm tired of dismissive anti-AI bias

When I was a kid, my family had an encyclopedia span an entire shelf in the living room's bookshelf. The shelf sagged under the weight, and the indexes overflowed onto the next shelf. One day, my father was at a friend's house and he saw a CD-ROM that claimed to hold an entire encyclopedia. He clicked around and found an entry titled "Space Shuttle" and watched a video of a rocket launching into space! As his eyes widened in amazement, he understood that he already bought his last physical encyclopedia.

Some time later, I remember a whole corner of the living room transformed into a computer desk. I don't know what happened to that shelf-spanning encyclopedia. In fairness, I don't know what happened to our Encyclopedia CD-ROM either.

I don't remember my first conversation with ChatGPT in December 2022, but at one point I asked it teach me the tax implications of short vs long-term capital gains in a rhyming scheme like The Night Before Christmas. As I read that poem describing the most mundane subject ever with proper rhymes and actually-decent tax advice, I realized this was my "watching a rocket launch into space" moment. 

I can't get behind dismissive attitudes about AI

I'm going to focus on the technology itself and not the cult of personality surrounding the people and companies working on the technology. I don't believe LLMs will directly lead to AGI. I'm also annoyed by the folks who hype it with the same passion as crypto bros. 

As new "thinking" techniques and agentic behavior takes off, I think LLMs will continue to incrementally improve and the real trick is finding ways to make them work with the known limitations they have. 

There's a few dismissive arguments against AI today.

"LLMs are just resource-hogging..."

Valid. AI crawlers are a real problem, and companies need to do better to be good stewards of the models they're creating. This is a solved problem with other types of crawlers on the internet and I expect to see better behavior after this initial gold rush is over.

On the flip side, user-initiated queries are an evolution of web search and they're amazing at reducing a 15 minute multi-tab Google search session into a thirty second Q&A. Like most other technology, efficiency will lead to continuous improvements in cost, environmental impact, and speed.

"...art-thieving..."

"It's stealing but also really cool" is a pretty concise way to state it. I see lots of ad hominem arguments for and against visual AI art, but they all devolve into specifics of the strawman that don't really apply. Using ChatGPT to print Mario posters and sell them on Etsy is already illegal. There's been attribution problems on the internet since the beginning.

It's a tool. As with every tool, it can be used to build things up or tear them down. Instead of attacking the tool, focus on the output from the tool. As always, you should support creators you like. I strongly support Penny Arcade, even if they don't care for AI either.

"...job destroyers."

Like many technologies before it, AI is just as capable of creating jobs as "destroying" them. Clocks put window-knockers out of a job. Cars put horseshit shovelers out of a job. Self-service gas pumps put gas station attendants out of a job (well, mostly). Each time, the standard of living rises for everybody. Software already automates the boring stuff and LLMs continue the tradition. "Regardless, 'don’t use power tools because they take people’s jobs' is not a winning strategy."

Don't worry, nobody's getting more free time!

"AI is just useless and wrong"

Oh no the Chinese are doing the useless thing that gives wrong answers cheaper
- John Scalzi on DeepSeek

This author wrote some of my favorite books, but I can't get behind this attitude. If you can't see the value of LLMs at all at this point, you're holding it wrong. 

Sure, LLMs hallucinate. Said cynically, they "lie." You need to add lots of context, or enable Web Search to ground the answers in external sources, or connect MCP servers, or or or... You can't just ask ChatGPT to "make me a fitness program," you need to enable Deep Research and give it links to existing workout programs, articles from Men's Health, detail your current workout plan, goals, previous injuries, and and and and...

This is LLM 101! Companies are starting to care about this skill. You can chalk this up as a UX problem, and knowing when to use LLMs takes practice. Often, you shouldn't even use them at all! I understand using the tools to understand the tools is an investment, but dismissing them outright means you miss out on any upside.

I'm using LLMs to build fun stuff and actually-useful stuff (how useful it actually is to get my blatherings in your inbox is up for debate). The best use cases for LLMs involve using them to help you achieve a result, not just copy/pasting their output and calling it a day. 

"AI is just making our kids dumb"

The kids that use ChatGPT to cheat on their homework would also cheat on their essays without ChatGPT. However, its simplicity is surely tempting kids who otherwise wouldn't cheat.

This isn't an AI problem, it's an education and parenting mission to improve how we teach our kids. We had to do the same thing when the internet came into existence. We had to adapt in similar ways when Wikipedia became prominent.

Professionals use these tools (internet, Wikipedia, and now LLMs) in our day jobs, and we all stand on the shoulders of giants to be as productive as we can. Kids absolutely need to learn fundamentals without outsourcing thinking to LLMs, but they don't learn it writing five-paragraph essays.

The challenge of education continues to be trying to inspire kids to be curious and care about learning new things. That's a hard problem, but it's never been a technology problem. It's an enduring educational battle. I'm usually inclined to believe the kids are gonna be alright.

LLMs are here to stay, until the next thing comes along

LLMs aren't the new crypto. You don't need to be technical to get all the benefits, but using LLMs with technical chops feels like a superpower right now. Simon Willison has a great guide on how he uses AI in his coding workflows which would immensely benefit software engineers everywhere. If you're technical, find excuses to use these tools!

I acknowledge I'm susceptible to Upton Sinclair's quote, "It is difficult to get a man to understand something when his salary depends on his not understanding it." I understand there's troublesome nuance to this technology, but I feel comfortable concluding that the pros are worth the cons. It's foolish to dismiss it entirely.