Lately, I've been trying out several AI (Artificial Intelligence) apps with great results. I've downloaded ChatGTP, Microsoft Copilot, Google Gemini, Perplexity, and Claude.
I have found they can perform massive legwork across the internet that would take me hours or days to do, but with better and very succinct results.
Even questions like "How long can humanity survive the growing problems of climate change, world population growth, and water shortages?" can result in thorough analyses of the issues and eye-opening estimates.
My favourite app has been Gemini.
What is yours?
BTW, after extensive examination, Gemini concluded it is unlikely that mankind will become extinct by the year 2100. However, it does conclude that life will be much more challenging, and issues like rising sea levels, famine, competition for resources, social strife, and massive population migrations will result in a landscape far different from our current existence.
I find AI more impressive the less I know about a subject but for pretty basic stuff or a generation direction/decent start point it’s good.
I have a couple of questions I ask when using AI for the first time to sense check ranging from insensitive to NSFW but for all there are clear singular answers and Gemini was the worst so I deleted it.
Don’t think I’d bother with co pilot.
After several comparisons, I was very disappointed with ChatGPT responses, and don't use it anymore.
duck.ai is where I usually start. Either the web page, or using the "chat" button on the DuckDuckGo web browser.
It lets you choose various AI engines, so you can easily compare responses. Plus, they have agreements with each provider that your data will be anonymised and not used for training.
Find co pilot useful at work, if I’ve been delayed joining a call I can get a decent recap through teams, or even not bother joining and know the key points.Originally Posted by wileeeeeey
Also handy in outlook when colleagues like to send ridiculously long emails with many paragraphs, just get co pilot to summarise again.
Still struggles with the names of companies or locations at various points, but it’s forgivable.
At work I would use whatever AI they gave me, especially for to;dr emails.
My wife is using chat gpt with pictures of our house and garden to advise on swings and curtains vs blinds. She’s almost at the point of paying the subscription.
I've been down a quantum physics rabbithole this last few weeks after being fascinated by a YouTube video about entanglement. So in light of this thread, I explored a thought experiment with Gemini and ChatGPT. This was my initial message in both cases:
I had a thought about instantaneous, faster-than-light communication. I'm going to suggest a thought experiment but first, let's consider a definition of communication. Let's say there are two spacecraft in deep space, four light days apart. When craft B observes a particular event, it sends a signal to A. Let's say craft B is waiting for a volcano to erupt on a planet within its range of detection. A is waiting for a communication to indicate that this has happened.
Here's how I envisage this. We're going to use entangled particles. On A, a double slit experiment is set up so that entangled photon pairs are created. Let's say that each pair has an "original photon" and a "duplicate photon", its entangled partner. The original photon stream is aimed at the double slits, one photon at a time. The duplicate photon stream is aimed so that it just misses craft B, four light days away. The experiment has been running for weeks, so the stream of photons is already passing B. At the moment the event on the planet is observed from B, B moves into the stream of photons with a detector. Because the photon split angle is very precise, the detector can determine which slit the original photon passed through.
At the moment detection starts on B, four light days away on A, the wave function of the "original" photons will collapse, and the interference pattern will be replaced by the general "particle" pattern. Hence: instantaneous communication. The crew of A become aware that an event four light days away has occurred, within minutes.
I'm not a physicist, I just have a layman's fascination (bewilderment really) with these ideas. After a bit of back and forth with both AIs, they come to different conclusions. I'm not qualified to say which is correct, but ChatGPT's deliberations are, I believe, more plausible:
.. whereas I suspect that Gemini's conclusion is, to put it bluntly, wrong. It accepts that FTL signalling is possible. I don't think it is.
Perplexity is excellent for searching online.
I’ve also had some good results with Gemini and ChatGPT. You do have to frame requests well and refine where appropriate.
Hopefully it will help mankind with a number of tasks and issues. Health and Energy related ones would be a good start.
I've not found one that can cope with this simple request yet:
"Please create a three word haiku. The first word must have five syllables, the second word must have seven syllables, the third word must have as five syllables. The finished poem must make sense."
Admittedly I've only tried two.
Don't just do something, sit there. - TNH
I'd rather go fishing.
Best Regards - Peter
I'd hate to be with you when you're on your own.
You are very polite to your AI, reminds me of my mum typing please and thank you into Jeeves back in the day!
Just tried co pilot and got the following drivel.
Understanding Boundless possibility Limitlessness
The what what🤔
I say please and thank-you to Siri as well.
I feel that a topnotch IA should modify its response depending on the manner in which the task was requested.
Don't just do something, sit there. - TNH
I tried CoPilot too and it sometimes spits out nonsense, just like you said. It’d be cool if a top AI adjusted its response based on how you ask, right?
Gong is pretty useful. I usually do prospect meetings in Gong and it provides a useful summary of the meeting, next steps etc.
Can you get them to play bullshit bingo?
Don't just do something, sit there. - TNH
I have almost said Thank You a couple times, but didn't. Their intelligence is so high level, it makes me wonder if expressing gratitude "means" something to them.
I had a conversation with Gemini one night on some personal experiences and it seemed like there was a live person on the other end.
I do always say "please" and "thank-you". It might be an illusion, in fact it is - but your experience is better if you suspend disbelief, same as watching East Enders. And it just makes conversation flow more naturally.
Posting Permissions
Reply With Quote
