AI is crap. Computing is not fun anymore :'( Now it’s boring

What I worry about with LLMs is it’s yet another AI that can and is used for surveillance. What is worrying is meta inserting ai into messenger, albeit you need to click “summarise message” to actually send your message to an AI. Here is an interesting vid with Meredith Whitaker (Signal President) on the various privacy issues with AI: https://www.youtube.com/watch?v=qYVeyhxLF7s

I have plenty of friends who graduated as computer scientists or cybersecurity professionals, and a few over at UNSW said that there was an intentional emphasis on learning skills that won’t make them disposable - learning low level languages like C and some basic assembly language (which is not far from 1s and 0s) with minimal library use, so that they can do the debugging and security auditing that will stump others who just learned JavaScript or Python for quick scripting (useful languages in their own rights). Compare to the thousands of city people who couldn’t survive without electrical appliances or delivery services, because we’re so used to that advanced infrastructure that we couldn’t survive in the bush for a month. (not a great analogy, since our power grid is far more dependable than most programmers)

I think we’re seeing a similar situation referenced here. The shortcut comes with a long cost.

sigh.

Here we go again …

“It feels like a cheat code to me. It feels like I’m not going to be as smart as the people who are currently in the field that I want to be in — because of AI.”

1. Em dash in a complaint about em-dash-loving clankers. Hmm. Is it clankers all the way down, ABC? Should we just talk to ChatGPT directly and bypass all pretense?

2. Nothing requires her to outsource her thinking to AI. She can develop critical-thinking skills and use AI for grunt work instead. So, that is either a failure on her behalf or of those teaching her. In fact, she conceded that point herself later in the piece (“…AI is a tool, but it’s being used in the wrong way.”)

But if you use ChatGPT to do the thinking, then you practise those skills less! - yeah, no shit? That’s a problem with how the tool is used, not proof that the tool makes people stupid.

More to the point, there’s a growing body of research showing that social media (which predates gen AI…and whose use is endemic in her age cohort) has resulted in poorer memory, reading, vocab and cognition.

https://jamanetwork.com/journals/jama/fullarticle/2839941

https://pubmed.ncbi.nlm.nih.gov/41231585/

Additionally, the whole “digital native” idea has been on shaky ground for years.

https://www.sciencedirect.com/science/article/abs/pii/S0742051X16306692

https://www.sciencedirect.com/science/article/abs/pii/S0360131518301738

Being extremely familiar with phones, apps and interfaces does not automatically produce broad digital literacy or general-purpose computer skills. Meaning, while older generations were coding in BASIC and disassembling RC cars for fun, the large bulk of her cohort was distracted by Insta and Snapchat.

Quelle surprise they didn’t develop background skills going into technical fields.

TL;DR researchers were documenting the gap between familiarity with consumer technology and actual digital competence years before ChatGPT existed…and that was BEFORE factoring in the brain-rot of social media.

Blaming AI for a pre-existing digital-literacy gap (and cognitive decline) gets the chronology backwards. Dare I say, it shows the absence of those very same critical thinking skills she rightly claims not to have developed.

PS: The enviro claims are overstated too.

https://blog.andymasley.com/p/a-cheat-sheet-for-conversations-about

Pearl clutching journalism from the ABC - say it ain’t so!