I wasn't sure where to commence my AI journey. What I am certain of, is the need to test the waters and what better way than to dive in and start swimming?
Right now, I am less worried about Image and Video (Multimodality) than day to day functionality. So I immediately ruled out Dall-e, Adobe Firefly, Luma, HeyGen et al.
The core models I considered were:
Confirmation bias & the mirror image (of ourselves).
I also did some asking around.
I felt women challenged the models both functionally and 'personally', (to include incorporating opinions and thoughts. I felt men challenged the models in more of a this <problem>, then <solution>. It does make me wonder about gender and confirmation bias - a balance I hope AI brings rather than polarises.
Open source
I am also part of the Google open source family. So I was a little bias to Gemini. On a more general query Gemini advised me that "Quantitative Risk Management by McNeil, Frey, and Embrechts" underpins my methodology. This is a real book - but I must confess I have not read it. I also doubt that I have as many inherent insights as the 3 authors (however I have taken the hint to add it to my reading list!)
The human in the middle.
When I went to university (a fun time), it was about the lecturers teaching us how to learn and to cite our references. We didn't have to know everything - but we had to know where to go to get the answers we needed.
I feel for the next generation, it will be about knowing how to prompt. And this is less easy than it sounds. You need to know what is known, and where the perimeter rests. Even Claude will give you a not fully informed opinion, if you don't feed in the focus on the relative weighting of the facts.
Where facts stop, hallucinations begin.
Disambiguation (and assumptions).
A friend was looking at a role in retail - Dune and Office are both familiar names on the UK High Street. I asked Claude for an opinion. This confused him.
"Futures/prediction markets — are you asking about how Dune and The Office traded on platforms like Kalshi, Polymarket, or the (now-defunct) Hollywood Stock Exchange?"
Dune the movie and Dune the shoe shop are quite different (I was glad that prediction markets was reflected back at me). Office the TV show is quite different to a South African owned footwear chain.
The role of Agency
Does agency come before intelligence?
Can AI be blackmailed?
Some excellent questions and reflections on OpenClaw by the excellent Dr Hannah Fry.
https://www.youtube.com/watch?v=WnzR5aOElvw
Fable 5.0
Yes, it's impressive. Before I knew it, I had worked with Fable for 5 hours. I was then timed out (for a 'smoko'). I guess even AI needs a break now and then.
So, anyway, for now, for me, the answer is Claude.
Hello Claude - Let's get at it!
I also did some asking around.
I felt women challenged the models both functionally and 'personally', (to include incorporating opinions and thoughts. I felt men challenged the models in more of a this <problem>, then <solution>. It does make me wonder about gender and confirmation bias - a balance I hope AI brings rather than polarises.
Open source
I am also part of the Google open source family. So I was a little bias to Gemini. On a more general query Gemini advised me that "Quantitative Risk Management by McNeil, Frey, and Embrechts" underpins my methodology. This is a real book - but I must confess I have not read it. I also doubt that I have as many inherent insights as the 3 authors (however I have taken the hint to add it to my reading list!)
The human in the middle.
When I went to university (a fun time), it was about the lecturers teaching us how to learn and to cite our references. We didn't have to know everything - but we had to know where to go to get the answers we needed.
I feel for the next generation, it will be about knowing how to prompt. And this is less easy than it sounds. You need to know what is known, and where the perimeter rests. Even Claude will give you a not fully informed opinion, if you don't feed in the focus on the relative weighting of the facts.
Where facts stop, hallucinations begin.
Disambiguation (and assumptions).
A friend was looking at a role in retail - Dune and Office are both familiar names on the UK High Street. I asked Claude for an opinion. This confused him.
"Futures/prediction markets — are you asking about how Dune and The Office traded on platforms like Kalshi, Polymarket, or the (now-defunct) Hollywood Stock Exchange?"
Dune the movie and Dune the shoe shop are quite different (I was glad that prediction markets was reflected back at me). Office the TV show is quite different to a South African owned footwear chain.
The role of Agency
Does agency come before intelligence?
Can AI be blackmailed?
Some excellent questions and reflections on OpenClaw by the excellent Dr Hannah Fry.
https://www.youtube.com/watch?v=WnzR5aOElvw
Fable 5.0
Yes, it's impressive. Before I knew it, I had worked with Fable for 5 hours. I was then timed out (for a 'smoko'). I guess even AI needs a break now and then.
So, anyway, for now, for me, the answer is Claude.
Hello Claude - Let's get at it!
No comments:
Post a Comment