Stephen Griffin
@stephengriffin.bsky.social
Rincewind-in-Residence, University of Birmingham https://www.stephengriffin.org/
Pinned
Have started my 8 year old off with a bit of 'vibe-coding', using Claude and ChatGPT to make games. The first - EXPEDITION: LEGACY OF THE FEATHERED FIENDS (yes, all caps) is now available to play! Link in comments ⬇️🐦🔥
Have started my 8 year old off with a bit of 'vibe-coding', using Claude and ChatGPT to make games. The first - EXPEDITION: LEGACY OF THE FEATHERED FIENDS (yes, all caps) is now available to play! Link in comments ⬇️🐦🔥
April 2, 2025 at 10:14 AM
Have started my 8 year old off with a bit of 'vibe-coding', using Claude and ChatGPT to make games. The first - EXPEDITION: LEGACY OF THE FEATHERED FIENDS (yes, all caps) is now available to play! Link in comments ⬇️🐦🔥
Perplexity now has an impressive Deep Research feature so I gave it the old Geoffrey vs Bungle test. Not too shabby.
www.perplexity.ai/search/who-w...
www.perplexity.ai/search/who-w...
Who would win in a fight between Geoffrey and Bungle from 1970s/80s itv...
The question of who would prevail in a physical confrontation between Geoffrey Hayes and Bungle, two iconic figures from the 1970s–80s British children’s...
www.perplexity.ai
February 14, 2025 at 10:15 PM
Perplexity now has an impressive Deep Research feature so I gave it the old Geoffrey vs Bungle test. Not too shabby.
www.perplexity.ai/search/who-w...
www.perplexity.ai/search/who-w...
January 28, 2025 at 7:49 PM
Reposted by Stephen Griffin
AI benchmarks falling into the McNamara Fallacy:
Step 1: Measure what can be easily measured
Step 2: Disregard that which cannot be measured easily
Step 3: Presume that which cannot be measured easily isn’t important
Step 4: Say that which can’t be easily measured doesn’t exist
Step 1: Measure what can be easily measured
Step 2: Disregard that which cannot be measured easily
Step 3: Presume that which cannot be measured easily isn’t important
Step 4: Say that which can’t be easily measured doesn’t exist
December 10, 2024 at 5:47 PM
AI benchmarks falling into the McNamara Fallacy:
Step 1: Measure what can be easily measured
Step 2: Disregard that which cannot be measured easily
Step 3: Presume that which cannot be measured easily isn’t important
Step 4: Say that which can’t be easily measured doesn’t exist
Step 1: Measure what can be easily measured
Step 2: Disregard that which cannot be measured easily
Step 3: Presume that which cannot be measured easily isn’t important
Step 4: Say that which can’t be easily measured doesn’t exist
Reposted by Stephen Griffin
Independent evaluations of OpenAI’s o3 suggest that it passed math & reasoning benchmarks that were previously considered far out of reach for AI including achieving a score on ARC-AGI that was associated with actually achieving AGI (though the creators of the benchmark don’t think it o3 is AGI)
December 20, 2024 at 6:26 PM
Independent evaluations of OpenAI’s o3 suggest that it passed math & reasoning benchmarks that were previously considered far out of reach for AI including achieving a score on ARC-AGI that was associated with actually achieving AGI (though the creators of the benchmark don’t think it o3 is AGI)
Reposted by Stephen Griffin
Sharks are older than the rings of Saturn.
This paper finds that the rings are no older than 400M years. Sharks date back to at least the Late Ordovician Period, 450M years ago.
This paper finds that the rings are no older than 400M years. Sharks date back to at least the Late Ordovician Period, 450M years ago.
December 8, 2024 at 6:54 AM
Sharks are older than the rings of Saturn.
This paper finds that the rings are no older than 400M years. Sharks date back to at least the Late Ordovician Period, 450M years ago.
This paper finds that the rings are no older than 400M years. Sharks date back to at least the Late Ordovician Period, 450M years ago.
youtube.com
December 4, 2024 at 9:16 PM
As the founding fathers once said, ‘Tesco had no turkey so we had to make do with chicken’. Happy Thanksgiving, America 🇺🇸
November 28, 2024 at 8:26 PM
As the founding fathers once said, ‘Tesco had no turkey so we had to make do with chicken’. Happy Thanksgiving, America 🇺🇸
Please rank 10 episodes of Star Trek TNG from Bouba to Kiki.
November 23, 2024 at 4:20 PM
Please rank 10 episodes of Star Trek TNG from Bouba to Kiki.
Reposted by Stephen Griffin
AI can help learning... when it isn't a crutch.
There are now multiple controlled experiments showing that students who use AI to get answers to problems hurts learning (even though they think they are learning), but that students who use well-promoted LLMs as a tutor perform better on tests.
There are now multiple controlled experiments showing that students who use AI to get answers to problems hurts learning (even though they think they are learning), but that students who use well-promoted LLMs as a tutor perform better on tests.
November 23, 2024 at 2:29 AM
AI can help learning... when it isn't a crutch.
There are now multiple controlled experiments showing that students who use AI to get answers to problems hurts learning (even though they think they are learning), but that students who use well-promoted LLMs as a tutor perform better on tests.
There are now multiple controlled experiments showing that students who use AI to get answers to problems hurts learning (even though they think they are learning), but that students who use well-promoted LLMs as a tutor perform better on tests.