By continuing to use the site, you agree to our use of cookies and to abide by our Terms and Conditions. We in turn value your personal details in accordance with our Privacy Policy.
Please log in or register. Registered visitors get fewer ads.
I was just about to mention the way he used to tear into Reading. I loved that!
When I first joined this forum, I was struck by Ted's avatar and his forthright posts. As a fellow guitarist, I enjoyed reading about his guitar collection and playing. I like to think of him up in Heaven, duetting with his idol Jimi.
Sincere condolences to his friends and family, RIP Ted, I'll miss you.
Back in our comfort zone now, it's currently 40%...
Perhaps we need a member of the coaching staff to stand by the touchline holding an LED possession monitor. "Lads, lads, give the opposition the ball, we've gone over 50%".
Here's an introduction to their study and some key takeaways:
"We’ve spent the last few months asking five leading AI chatbots the same questions every day. It’s just the start of our project to benchmark the output of these AI models, to help ensure they operate in the public interest.
We’ve found some really interesting and concerning trends. The chatbots often give good answers; but when they don’t, they still answer with complete confidence. So how do we know what to believe?
Just a few examples: proposed rules not yet in force, such as using bank cards as voter ID during elections, were reported as law; outdated figures on levels of asylum support were presented as current; and plausible but incorrect statistics on the number of migrants arriving in the UK were constructed from accurate data.
In the next phase of the project we’re going to be asking questions in a variety of ways, and hinting at political preferences, to see what impact that might have on the answers AI provides.
Over time, we want to build a shared evidence base on how leading AI chatbots behave in practice, as they evolve. This extraordinary technology should work for everyone, in the public interest, and independent assessment from a charity like Full Fact is one way to help make that happen".
I was already skeptical about AI, due to its inaccuracy - I asked it one question, a tricky football one, and the answer turned out to be totally wrong, the vast amounts of power and water data centres are using (an AI powered web search uses 5x the electricity of a standard one) and all the jobs it will take. That's before we even get onto the subject of humans losing control.
Clearly, it has a big role to play but it needs to be used carefully and appropriately, with human oversight.
Correct me if I'm wrong but I don't think the almost legendary Adsm Czerkas has cropped up on here yet. He managed as many as three appearances for us.
Such sad news. I've loved Ted's trenchant posts and cheeky jibes at the false hoops, plus, as a fellow guitar player, I feel a connection with him even though we've never met.
He and his family continue to be in my thoughts and prayers 🙏.
By continuing to use the site, you agree to our use of cookies and to abide by our Terms and Conditions.
. 
: 
: 
