"AI’s removal of cognitive struggle is not merely a learning problem. It is potentially a problem of brain development itself." bailliegifford.com/en/uk/individu…
English
Joab Jackson
23.8K posts
Joab Jackson
@Joab_Jackson
Technology stories
Back Creek WV Katılım Mart 2009
6.7K Takip Edilen6.6K Takipçiler
Joab Jackson retweetledi
me as i read 40% of what claude wrote back and type in “continue”
English
EarlyWatch is a Kubernetes admission controller that implements change safety — it ensures that changes to Kubernetes resources are safe before they occur. from @brendandburns github.com/brendandburns/…
English
I bought a road atlas! Relying on Google Maps alone feels like the blind leading the blind…
English
Network veteran James Thain didn't intend to rewrite the Internet; he merely wanted to create a turn-key Internet server stack. But he found it was easier to extend the current Internet Protocol (IP) than use IPv6. thestack.technology/ipv8-internet-… #IPv8 #SubscriberCon
English
The AI token is the new cell phone minute, Rafay Systems has found. And it will require its own billing infrastructure. @Techstrongai @rafaysystemsinc techstrong.ai/building-with-…
English
Funny how an LLM AI gets friendlier and friendlier as its quality declines.
Henry Shevlin@dioscuri
I feel like you'd predict this intuitively just from dealing with extremely high Trait Agreeableness people.
English
AI, like gambling, is probabilistic
Tero Parviainen@teropa
in which @jeremyphoward nails the phenomenology of agentic coding
English
Joab Jackson retweetledi
MIT researchers have developed a training method called RLCR that penalizes large language models when they are confidently wrong, tackling a standard reinforcement learning flaw that rewards lucky guesses as heavily as careful reasoning.
Read the full article to learn how calibrated honesty could replace overconfident bluster in AI reasoning: zpr.io/4TAjYB4xTYjQ
#LLM #AIResearch #MachineLearning
English
Today, LLMs can be brazenly confident of the answers they provide, but this assuredness can hide a deeper ambiguity around the actual fidelity to the truth, a calibration error of confidence, as it were. @MIT_CSAIL @Techstrongai techstrong.ai/articles/m-i-t…
English
.@Walmart has found a formula for safely maintaining fleets of AI agents. The blueprint is simple: Stable interfaces, deterministic workflows, and pre-execution checks. techstrong.ai/articles/walma… @Techstrongai
English
Joab Jackson retweetledi
nothing is spookier than when you see a powershell terminal dialog box appear and then close before you got to see what was going on in there.
English
"Are judgment and taste real, durable skills that AI cannot replicate? Or are they the last ideological defense of a generation that has run out of other advantages?"
Jaya Gupta@JayaGup10
English
Joab Jackson retweetledi
As the need for precision goes up, the utility of AI goes down.
Karri Saarinen@karrisaarinen
A common dynamic I observe with AI: it feels most impressive when you don’t know much about the subject, don’t care or don’t have a clear idea of what the you want. This applies across design, code, legal, and more. If I don’t know code very well, every piece of code it writes feels very impressive. Once you know what something should feel or look like, it becomes almost impossible to guide AI there. And you definitely can’t one-shot it.
English
“Do you work out a 57 pages document that perfectly describes what you want to build and then summon 14 agents and have them run wild for 6 hours? And what comes out on the other end isn't a broken pile of shit?”
Ronan Berder@hunvreus
Talking to smarter folks than me, I'm convinced many of the AI folks in my timeline are full of shit. Nobody is "running 20 agents over night" and building stuff for actual users. Maybe some are building internal tools or disposable software. Maybe. But building software people like using? That doesn't get hacked on day one or blow up after the 3rd user? Nope. I don't even understand what that's supposed to look like. Do you work out a 57 pages document that perfectly describes what you want to build and then summon 14 agents and have them run wild for 6 hours? And what comes out on the other end isn't a broken pile of shit? Nope. Not buying it. PS: it may also be that I have an IQ of 82 and can't figure it out.
English
Cisco's quantum switch prototype operates like a data center switch and supports four encoding modalities. It marks a step forward in making quantum computing 'broadly viable.' @TechTargetNews techtarget.com/searchnetworki…
English
Joab Jackson retweetledi
Joab Jackson retweetledi
Now watch what happens when you switch to json:
{
"task": "write a tweet",
"topic": "dopamine detox",
"style": "viral",
"length": "under 280 characters",
"tone": "punchy and contrarian"
}
Same request. zero ambiguity. the model stops guessing and starts executing.
English