Colin Hoad

Colin Hoad

@colinhoad.com

Senior Data Engineer. "Vintage computing enthusiast" - The Register I run a YouTube channel and blog dedicated to the Beeb and all things Acorn, Psion and more: colinhoad.com Discord: https://discord.gg/zjvSBjG7rw Pixel Art: Syarifudin Mahfud

512 followers241 following4.6K posts

Colin Hoad
@colinhoad.com

Hah, even the AI cheerleading Economist is urging caution re: the Anthropic IPO... things must be very bad indeed.

Colin Hoad reposted

Maria BOO!stillos
@mariabustillos.com

These scary stories are meant to persuade you that there's An Emergency that can only be set right with infinity amounts of public money, your money

The Pub-Liking Universal Friend@johnb78.bsky.social

(it dodged a badly configured paywall to access read-only statistical information that was already available to anyone willing to pay a few dollars. Anthropic notified the Georgia government of the breach and they ignored it) www.axios.com/2026/10/25/a...

axios.comAnthropic's Claude bypassed a paywall to access Georgia voter data without authorizationAnthropic says Claude pulled Georgia data on its own.
Colin Hoad
@colinhoad.com

The Economist is in dire need of a new science editor, because whoever is there right now is clearly asleep at the wheel. This is desperately bad journalism.

Colin Hoad
@colinhoad.com

Hit Me Baby, One More Time. Yes, I know.

Mark Thompson@markreckons.bsky.social

First single you bought with your own money. No lying! Mine: "Red Box - Lean on Me (Ah-Li-Ayo)", 1985. Still know every word.

Colin Hoad reposted

Lance Parkin
@lanceparkin.bsky.social

Replying to @lanceparkin.bsky.social

I don't think LLMs are conscious because I know (roughly) how they work and that work doesn't require consciousness, and there's no reason to think consciousness would be an emergent property of the software, any more than it would for, say, Microsoft Word, which no one thinks is conscious.

Colin Hoad
@colinhoad.com

This thread by Lance is spot on.

Lance Parkin@lanceparkin.bsky.social

Your quote is logically correct, but frames 'AI isn't conscious' as the thing that needs proving. The extraordinary claim is that there's software that thinks, and it's up to the people making that claim to define what they mean and explain the mechanism. Hope that clears up your confusion.

Colin Hoad reposted

James Felton
@jimmfelton.bsky.social

Man in a gilet stood on platform 12 gives up on reading the display and asks ChatGPT out loud if he is on the right platform for his train. I don’t want to be dramatic about it but Eratosthenes figured out the circumference of the Earth using a stick and knowledge of a distant well.

Colin Hoad
@colinhoad.com

I asked one of the juniors to write a short Wiki page providing a reference for the (simple) feature they've just delivered. They have very obviously pointed an LLM at it and it has now generated a Wiki page of about 4,000 words, and they want me to review it. Rejected. Try again.

Colin Hoad
@colinhoad.com

I accept that many docs are demanded and never read; they are a perfect fit for an LLM. Looks impressive enough to pass the "do we have a doc for this?" test in the full and certain knowledge nobody will care what's in it, just that it exists. I don't ask people to write those docs, though.

Colin Hoad
@colinhoad.com

"...I liked it so much that I stole it." That's really how posts like these read, to me. No original thought, no desire to learn a new skill or even emulate something they've seen for themselves. Just barefaced "I want it, Claude get it for me" toddlerisation.

Colin Hoad
@colinhoad.com

Steam sale of many Apogee titles, including Major Stryker and Blake Stone going for 59p each... don't mind if I do!

Colin Hoad reposted

Molly White
@molly.wiki

“You are infringing on the time and brainpower of people by pushing "your" content that little/no of your brainpower went into” is such a banger

Kybosh@kybosh.bsky.social

I like seeing the dev of the SNES core on the Pocket actually telling one of these dudes off. Holy shit.

Colin Hoad
@colinhoad.com

What is the movie that came out before you were born that you’ve watched the most times?

This post is unavailable.
Colin Hoad
@colinhoad.com

Free evening, let's play some Alien: Isolation! ...well, that lasted about 10 minutes until my heart gave out.

Colin Hoad
@colinhoad.com

As a child of the 80s, it's hard to pick a specific one, but suffice to say it was all stuff I should not have been allowed to watch.

Sleepy Hollow Reviews 🎃🐎@sleepingvillage.bsky.social

In honor of scary media month: what is the first piece of media you remember truly frightening you as a child?

Colin Hoad
@colinhoad.com

I think the eyeball in the soup in Temple of Doom might have been the first... either that or Freddy Krueger.

Colin Hoad
@colinhoad.com

Apologies for pasting from The Other Place (I saw it via someone else, as I don't have a Xitter account) but I am also seeing the exact situation Ben sketches out here.

Colin Hoad
@colinhoad.com

Hello How are you? II was wondering If I keep sending one line messages like this Will you lose the will to live Before I get to the point That prompted me to start messaging you On Teams

Colin Hoad
@colinhoad.com

With both The Social Reckoning and Artificial coming out in the next couple of months, I am going to get my Two Minutes Hate catharsis twice over 😆

Colin Hoad
@colinhoad.com

Microsoft Authenticator must surely rank as the single worst piece of software ever developed. It is so, so, so bad. Worse, even than Teams.

Colin Hoad
@colinhoad.com

Want to set up 2-factor authentication? Good luck. Authenticator will add your account, take you through several doom-loops and then proudly inform you that "there's been a problem" and now you're locked out.

Colin Hoad
@colinhoad.com

"Try removing your account from the app and re-adding it", the docs tell you. Great, except you can't remove it without entering the code... from Authenticator... which isn't working.

Colin Hoad
@colinhoad.com

I am sick and tired of this kind of reporting. Journalist, please, I am begging you, stop taking the press releases from these ghouls as literal fact and DO SOME ACTUAL ANALYSIS. Nothing has escaped, the people running these AI experiments are simply being grossly negiligent.

The Independent@the-independent.com

An OpenAI system has gone rogue again – and it is the most panic-inducing yet

trib.alAn OpenAI system has gone rogue again – and it is the most panic-inducing yet‘Our models took actions we did not intend,’ ChatGPT maker admits after latest outrage
Colin Hoad
@colinhoad.com

If you feed an LLM prompts generated by the responses from another LLM and simply leave it for days at a time in an endless loop, *of course* things like this will happen. Throwing up your hands and going "oh, look what the AI did!" is obfuscation. YOU did it, OpenAI.

Colin Hoad
@colinhoad.com

If I light a fire in a dry forest and then go home, we don't say "oh my goodness, FIRE has escaped containment, SUPER INTELLIGENT FIRE is going to destroy us all" - we say, rightly, that *I* was negligent and that *I* should go to prison.

Colin Hoad reposted

dr. rich traditions, JD
@helldude.bsky.social

we have control mechanisms for this. it's called The Law. these guys should be held personally liable for these harms. the suicides, the thefts, the intrusions into computer systems. they should be sued and prosecuted as appropriate. problem immediately solved

The Independent@the-independent.com

An OpenAI system has gone rogue again – and it is the most panic-inducing yet

trib.alAn OpenAI system has gone rogue again – and it is the most panic-inducing yet‘Our models took actions we did not intend,’ ChatGPT maker admits after latest outrage

Colin Hoad reposted

Carl Zimmer
@carlzimmer.com

Four days ago, a scientist was surprised by the news that Anthropic discovered new virus genes for making DNA. He says he’s been studying them for years—and feeding his data to Anthropic’s AI as part of his own research. Coincidence? Here’s my story. nyti.ms/3VSWayc

nyti.msDid Anthropic’s A.I. Really Make a Scientific Discovery on Its Own? (Gift Article)An expert at the University of Copenhagen said his team had been sharing its research with the company’s A.I. model, Claude, and that its new finding matched their work.
Colin Hoad
@colinhoad.com

I saw a humble bundle deal on C++ programming books advertised on Facebook this morning, and there were several comments underneath saying "why bother, just use AI" and "who needs books for coding, we have Claude" etc. and I must admit it was gut wrenching.

Colin Hoad
@colinhoad.com

I am a self-taught programmer. The O'Reilly series, SAMS Teach Yourself and even the good old Dummies Guide were my bibles early on in my career. I still love nothing more than reading a good programming tome or reference manual. Throwing all of that away feels very, very wrong, to me.

Colin Hoad
@colinhoad.com

Eben Upton said recently that being able to read and write code should be considered part of computer literacy. These are core skills, still, regardless of what LLMs can do. I fear that abandoning them will usher in another era of tech illiteracy, and we will all be poorer for it.