@gianluca.mauro: The AI graph that made everybody go nuts #Ai #LearnOnTikTok #artificialintelligence #business #machinelearning #product #ux #entrepreneur

Gianluca Mauro
Gianluca Mauro
Open In TikTok:
Region: DK
Friday 03 January 2025 14:36:24 GMT
11162
442
22
15

Music

Download

Comments

thanossofroniou
Thanos Sofroniou :
The only "useful" application of AI technology was Alphafold by deepmind. But o3 is a much bigger leap than Alphafold
2025-01-03 14:57:53
7
robk98
Kelflcwlalwo :
Are they maybe specifically training on these kinds of tasks with the new models?
2025-01-03 14:53:39
2
altered_realms
Altered Realms (AR) :
similar to IQ test
2025-01-06 12:02:11
0
alexigneel
alexigneel :
is IT comparable to IQ Tests?
2025-01-03 15:01:12
0
phasn
Hasn :
true
2025-01-05 08:35:58
0
nummer9527
编号9527 :
Until the model can self explain its logic chain aka the reasoning process, I wont call it AGI. There is clear line between logic inference and statistical inference.
2025-01-25 08:51:38
1
mr_brown2
Mr_Brown :
It's just like iq tests
2025-01-03 14:44:04
0
philkoshy
Philip Koshy442 :
o3 also scored over 70% on swe-bench. They are on track to saturate all benchmarks...
2025-01-03 22:02:49
0
_k1ngcr0wn
_k1ngcr0wn :
Guessing the next sequence in a pattern is so much of what AI does.
2025-01-03 16:47:04
0
felipepkorbes
FelipePKorbes :
if AI look at this test and finding that there is only one acceptable interpretation, then it is lacking in creativity, wich is pretty essential to AGI IMO. the thinking process is important
2025-01-03 19:04:09
0
profmata
Ricardo Mata :
it looks like pattern recognition, which is something a good AI should excel at, but gives me little information for AGI imho
2025-01-04 07:49:24
0
aiscream.de
Sascha | AIscream :
It's coming. But not in 2025. 😂
2025-01-03 21:19:13
0
hoops_55
Steve :
If standard aptitude tests are the benchmark, that’s not exactly super intelligence really?
2025-01-03 14:56:43
0
__zaak__
__Zaak__ :
Where does Deep Seek AI v3 compare to better ones on the graphic … it’s China’s 🇨🇳 Open Source model
2025-01-03 18:02:10
0
sebastian93860
user5219523264888 :
I had o4 write me a parser in Common Lisp but it made mistakes and it could not „understand“ my explanations what was wrong. It changed the specification after a few tries
2025-01-03 16:27:27
1
michaelgerges27
Michael Gerges918 :
It’s similar to human iq tests. Now we have a program outperforming Drs lawyers, marketers and administrators all together and can do well what we have defined as IQ, what else do you want?
2025-01-04 02:26:32
0
To see more videos from user @gianluca.mauro, please go to the Tikwm homepage.

Other Videos


About