Hi there
While I Iike AI it still has a long way to go yet - for example trying to use say chatGPT to return the market capitilisation of the current UK's AIM market (for those who aren't into finance this is the UK's "Alternative Investment Market" - for companies far smaller than would be traded on the standard markets such as S&P (NYSE) , NASDAQ, FTSE 100/250 etc) - usually also very risky but can once in a while return a winner. - Spread can be large (i.e difference between offer and bid price), plus selling quickly might be difficult or impossible as there's not always a ready market.
For example about 4 mins to return this:
Splitting AIM companies by market cap size :
£0–£50M: ~300 companies
£51M–£100M: ~150 companies
£101M–£150M: ~80 companies
£151M–£200M: ~50 companies
£201M–£250M: ~35 companies
£251M–£300M: ~25 companies
£301M–£350M: ~20 companies
£351M–£400M: ~15 companies
£401M–£450M: ~7 companies
£451M–£500M: ~3 companies
it took quite a while to come up with (admittedly the right set of data) but simply used brute force rather than get a load of the required data from various financial sites which had a load of this data already available. Especially on decent hardware and with excellent Internet at around 10 Gbps.
Anybody got any insite as to the "efficiency" of the learning processes of some of these A.I systems or is that going to be forever a "closed subject" !!!!. I believe python was a computer language primarily designed for use with robotics and A.I.
cheers
jimbo
While I Iike AI it still has a long way to go yet - for example trying to use say chatGPT to return the market capitilisation of the current UK's AIM market (for those who aren't into finance this is the UK's "Alternative Investment Market" - for companies far smaller than would be traded on the standard markets such as S&P (NYSE) , NASDAQ, FTSE 100/250 etc) - usually also very risky but can once in a while return a winner. - Spread can be large (i.e difference between offer and bid price), plus selling quickly might be difficult or impossible as there's not always a ready market.
For example about 4 mins to return this:
Splitting AIM companies by market cap size :
£0–£50M: ~300 companies
£51M–£100M: ~150 companies
£101M–£150M: ~80 companies
£151M–£200M: ~50 companies
£201M–£250M: ~35 companies
£251M–£300M: ~25 companies
£301M–£350M: ~20 companies
£351M–£400M: ~15 companies
£401M–£450M: ~7 companies
£451M–£500M: ~3 companies
it took quite a while to come up with (admittedly the right set of data) but simply used brute force rather than get a load of the required data from various financial sites which had a load of this data already available. Especially on decent hardware and with excellent Internet at around 10 Gbps.
Anybody got any insite as to the "efficiency" of the learning processes of some of these A.I systems or is that going to be forever a "closed subject" !!!!. I believe python was a computer language primarily designed for use with robotics and A.I.
cheers
jimbo
Last edited:
My Computer
System One
-
- OS
- Windows XP,7,10,11 Linux Arch Linux
- Computer type
- PC/Desktop
- CPU
- 2 X Intel i7
