THE major AI player hasn’t released agent software. Do they know something we should?
Plus 5 Cases that depict the loss of morality technology brings, an update on one of my 2025 predictions, and Headline Head Scratchers
MAIN COURSE: Can we learn about digital agents gone mad from animation?
Many organizations have announced their digital agents, Microsoft and Anthropic of note, and of course the literal trumpeting that I covered in earlier Tonics by the marketing baboon that is Salesforce. So one wonders why OpenAI, who started research in this area very early, has never announced their offering. It’s not like thy are the shy ones in the digital jungle.
Digital agents are AI entities (the first step in being declared a legal person is to move from just being software to being referred to as an entity) capable of interacting with their environment, in order to to autonomously complete tasks without human intervention. A simple definition, but with possible wild outcomes.
Back to OpenAI. What is it that they know? Turns out that they are worried about how easily it is to get AI software to ignore instructions and jump over guardrails (see the bomb reference from last week). So suddenly they are worried about a hacker getting control of customers’s computers but not the same people downloading bomb-making instructions? My cynical self smells lawsuits in the former. Think about what might happen to these agents when they are integrated with robots.
Well the BBC has already done an excellent documentary on this very subject. As of this month it has been released on Netflix globally so I invite everyone to watch it. It is in a unique animated format but the brilliant creative people at Aardman Animations gnow clearly what trouble we are all in for. I refer to the Wallace and Gromit epic Vengeance Most Fowl (I lied, it is not a documentary). The story focuses on an AI driven robotic gnome that does gardening, being hacked by a criminal and changed into an army of evil doing robots. I encourage you to watch it. Seriously.
SECONDI: 5 Cases that depict the loss of morality technology brings
We will start this week with Open AI at it’s normal mercenary best - which makes my main course above so mysterious - doing pretty much whatever they want while burning through untold billions. Several different reports of what I will call its ingestion spider (their bot software that goes through different websites absorbing everything it can into OpenAI’s training environment, copyright be damned) causing havoc with corporate websites. Here is a case of them bringing a small 7 person digital business to its knees by repeatedly scraping their dense and rich website that slowed it to a crawl like it was a denial of service attack.
Evidence from a legal case against Meta, a perpetual immoral corporate actor, was released last week showing them to be blindly scooping up pirated copyrighted material without a worry in the world. Internal email evidence shows that approval to do so went right to the top (signed MZ). They need to get some ex-politicians onboard quickly to show them how to have private personal emails, and permanent organizational email deletion policies in place to avoid culpability and accountability.
On a much gloomier note many of you may have heard about the suicide of a 14 year old when interacting with character.ai. That is the same website I highlighted last week where a musician was clowning around with himself though many of the exchanges were about violence, always a laughing matter. I have no idea why we continue to think it is a good idea to make software so engaging with realistic mimicry without have limitations or restrictions in place, just to make money. Then we have to rally ourselves after horrifc incidents like this to get proper regulations put into place. Like the thanks we need to show to Jon Haidt for dedicating about 18 months of his life to get cel phone usage in schools mostly stopped and education back on track. Hopefully people like Jon emerge soon to get appropriate AI regulation and governance in place
But it isn’t just businesses that engage in dubious behaviour. Here is a story reporting a civilian crashing their drone into a crucial firefighting plane over the wild Palisades fire in the LA area. The plane managed to land safely but was damaged significantly. Officials are attempting to track down the person responsible. We know what the drone was doing. It was getting on the spot, up close and personal videos of the fire in order to satisfy the literal tens of thousands ghouls glued to their favorite social media platforms demanding the absolute latest in visuals of this disaster. To me, they are all responsible.
Another story of a person acting poorly at the LA fire that shows the pernicious and infectious effects of technology. The quote below is from The Free Press describing the types of human behaviour that is repugnant to me and is one of the reasons I started writing this publication. Bolding added.
“A few hours later I saw there were deadly fires in the Palisades, some 25 miles away, and felt badly for posting fun videos on Twitter, though not badly enough to take them down, since they were getting a lot of likes”
In karmic justice their house also burned. Unfortunately I don’t believe in nor can we rely on karma to protect us.
ENCORE: Checking one of my 2025 Predictions
I am taking my 10 predictions for 2025 somewhat seriously and am tracking them closely. The latest court case from Friday suggests that TikTok is in poor legal shape at the Supreme Court to try to escape the legal order for their parent company to divest them. But the Whack-a-do soon to be in the large white house has yet to be heard from. He seems somehow to be able to slither around the law.
HEADLINE HEAD SCRATCHERS
How can something that has no physical presence or taste buds create something better than Double Stuf?
We are getting closer to peak AI hype and further away from grounded reality.
That’s a wrap until Thursday. I appreciate everyone that reads some or all of this. As always comments or sharing welcome