Shortly before the pandemic when my son was 18, one of his group had a reputation among the girls as someone who enjoyed imposing his enthusiasm for choking on his partners.
None of them really know where he picked it up from.
All of them knew not to encourage it (or him)
None of them really know where he picked it up from.
All of them knew not to encourage it (or him)
July 7, 2025 at 11:00 PM
Shortly before the pandemic when my son was 18, one of his group had a reputation among the girls as someone who enjoyed imposing his enthusiasm for choking on his partners.
None of them really know where he picked it up from.
All of them knew not to encourage it (or him)
None of them really know where he picked it up from.
All of them knew not to encourage it (or him)
No. Nope. Not happening.
Toe separatists and their instruments of partition are among the great evils of our age.
Society was too passive when the flip flop and it's innocent little bar between the first 2 toes came along.
We must resist. Digiti pedis Unite!
Toe separatists and their instruments of partition are among the great evils of our age.
Society was too passive when the flip flop and it's innocent little bar between the first 2 toes came along.
We must resist. Digiti pedis Unite!
July 7, 2025 at 10:52 PM
No. Nope. Not happening.
Toe separatists and their instruments of partition are among the great evils of our age.
Society was too passive when the flip flop and it's innocent little bar between the first 2 toes came along.
We must resist. Digiti pedis Unite!
Toe separatists and their instruments of partition are among the great evils of our age.
Society was too passive when the flip flop and it's innocent little bar between the first 2 toes came along.
We must resist. Digiti pedis Unite!
Give them something to vote for, not just something to vote against.
Beautiful philosophy. Everyone wins
Beautiful philosophy. Everyone wins
July 1, 2025 at 9:41 PM
Give them something to vote for, not just something to vote against.
Beautiful philosophy. Everyone wins
Beautiful philosophy. Everyone wins
First we thought we could make something useful.
Now we've made it, we see we have no idea how to make it do anything useful. And that might bankrupt us.
So we're going to focus on making it profitable. For us.
And let other schmucks see if they can make it useful before it bankrupts them
Good luck
Now we've made it, we see we have no idea how to make it do anything useful. And that might bankrupt us.
So we're going to focus on making it profitable. For us.
And let other schmucks see if they can make it useful before it bankrupts them
Good luck
June 30, 2025 at 5:01 PM
First we thought we could make something useful.
Now we've made it, we see we have no idea how to make it do anything useful. And that might bankrupt us.
So we're going to focus on making it profitable. For us.
And let other schmucks see if they can make it useful before it bankrupts them
Good luck
Now we've made it, we see we have no idea how to make it do anything useful. And that might bankrupt us.
So we're going to focus on making it profitable. For us.
And let other schmucks see if they can make it useful before it bankrupts them
Good luck
Tell me you've never bought a Brabantia without telling me
June 28, 2025 at 12:05 PM
Tell me you've never bought a Brabantia without telling me
This may not quite be too your point but having named our dog after a character from Discworld, I'm always surprised how few people have heard of Discworld or Pratchett.
Doubly so because we live 5 miles from where he was born.
Doubly so because we live 5 miles from where he was born.
June 28, 2025 at 12:03 PM
This may not quite be too your point but having named our dog after a character from Discworld, I'm always surprised how few people have heard of Discworld or Pratchett.
Doubly so because we live 5 miles from where he was born.
Doubly so because we live 5 miles from where he was born.
Take a look at Nothing phones
June 22, 2025 at 8:32 PM
Take a look at Nothing phones
That's not Norfolk, it's Hunstanton.
A totally justifiable misunderstanding if we're being honest.
A totally justifiable misunderstanding if we're being honest.
June 20, 2025 at 7:19 PM
That's not Norfolk, it's Hunstanton.
A totally justifiable misunderstanding if we're being honest.
A totally justifiable misunderstanding if we're being honest.
Can we bring back chemistry puns too?
Or would that be overreacting?
Or would that be overreacting?
June 20, 2025 at 7:07 PM
Can we bring back chemistry puns too?
Or would that be overreacting?
Or would that be overreacting?
Indeed.
The greatest issue, and what @spavel.bsky.social is calling out, is the gulf between academic and business expectations for LLMs. And how the business expectation is magically translating into $billions in investment.
Just like crypto, the right level will eventually be found
The greatest issue, and what @spavel.bsky.social is calling out, is the gulf between academic and business expectations for LLMs. And how the business expectation is magically translating into $billions in investment.
Just like crypto, the right level will eventually be found
June 20, 2025 at 6:06 PM
Indeed.
The greatest issue, and what @spavel.bsky.social is calling out, is the gulf between academic and business expectations for LLMs. And how the business expectation is magically translating into $billions in investment.
Just like crypto, the right level will eventually be found
The greatest issue, and what @spavel.bsky.social is calling out, is the gulf between academic and business expectations for LLMs. And how the business expectation is magically translating into $billions in investment.
Just like crypto, the right level will eventually be found
We live in very stupid times because plenty of people in business take a 25% success rate at mathematical tasks as sufficient evidence that LLMs are now ready to be put to any and every business task, with minimal use training.
That's more a failure of business & marketing than maths or computing
That's more a failure of business & marketing than maths or computing
June 20, 2025 at 3:33 PM
We live in very stupid times because plenty of people in business take a 25% success rate at mathematical tasks as sufficient evidence that LLMs are now ready to be put to any and every business task, with minimal use training.
That's more a failure of business & marketing than maths or computing
That's more a failure of business & marketing than maths or computing
Cynically this probably explains why marketing use cases seem to be a focus.
(Joking)
(Joking)
June 20, 2025 at 3:12 PM
Cynically this probably explains why marketing use cases seem to be a focus.
(Joking)
(Joking)
They operate at the level of a 9 year old child.
Can't tell fact from opinion or cite sources & make stuff up for extra credit.
Question to any CEO - which parts of your business are you happy handing over to 9 year old kids?
Can't tell fact from opinion or cite sources & make stuff up for extra credit.
Question to any CEO - which parts of your business are you happy handing over to 9 year old kids?
June 20, 2025 at 3:11 PM
They operate at the level of a 9 year old child.
Can't tell fact from opinion or cite sources & make stuff up for extra credit.
Question to any CEO - which parts of your business are you happy handing over to 9 year old kids?
Can't tell fact from opinion or cite sources & make stuff up for extra credit.
Question to any CEO - which parts of your business are you happy handing over to 9 year old kids?
It certainly is. How's the job hunt going?
Or is that not a question friends ask
Or is that not a question friends ask
June 19, 2025 at 9:56 PM
It certainly is. How's the job hunt going?
Or is that not a question friends ask
Or is that not a question friends ask
Possible I'm conflating then with agents, which do follow rules/process design but with some fuzzy logic
June 19, 2025 at 6:09 PM
Possible I'm conflating then with agents, which do follow rules/process design but with some fuzzy logic
I get that.
I'm just bemused that "a computer program" apparently can't consistently follow rules.
In this situation, use the structure data.
In that situation, ignore it.
Feels like there's a layer of hubris. People so keen to show it being autonomous and intelligent they dare not give it rules
I'm just bemused that "a computer program" apparently can't consistently follow rules.
In this situation, use the structure data.
In that situation, ignore it.
Feels like there's a layer of hubris. People so keen to show it being autonomous and intelligent they dare not give it rules
June 19, 2025 at 5:16 PM
I get that.
I'm just bemused that "a computer program" apparently can't consistently follow rules.
In this situation, use the structure data.
In that situation, ignore it.
Feels like there's a layer of hubris. People so keen to show it being autonomous and intelligent they dare not give it rules
I'm just bemused that "a computer program" apparently can't consistently follow rules.
In this situation, use the structure data.
In that situation, ignore it.
Feels like there's a layer of hubris. People so keen to show it being autonomous and intelligent they dare not give it rules
Yes, I've just found @drkatedevlin.bsky.social and immediately followed.
I am having a big sit down while reconciling myself with the difference between academic and business discourse around AI & LLMs
Appreciate your responses Saeed.
I am having a big sit down while reconciling myself with the difference between academic and business discourse around AI & LLMs
Appreciate your responses Saeed.
June 19, 2025 at 4:53 PM
Yes, I've just found @drkatedevlin.bsky.social and immediately followed.
I am having a big sit down while reconciling myself with the difference between academic and business discourse around AI & LLMs
Appreciate your responses Saeed.
I am having a big sit down while reconciling myself with the difference between academic and business discourse around AI & LLMs
Appreciate your responses Saeed.
This feels reasonable to me.
Children should learn about the shallow, duplicitous, mind-numbing realities of commerce and industry as places of work in a gentler, more humane way than by being exposed to undiluted LinkedIn
Children should learn about the shallow, duplicitous, mind-numbing realities of commerce and industry as places of work in a gentler, more humane way than by being exposed to undiluted LinkedIn
June 19, 2025 at 3:01 PM
This feels reasonable to me.
Children should learn about the shallow, duplicitous, mind-numbing realities of commerce and industry as places of work in a gentler, more humane way than by being exposed to undiluted LinkedIn
Children should learn about the shallow, duplicitous, mind-numbing realities of commerce and industry as places of work in a gentler, more humane way than by being exposed to undiluted LinkedIn
What confuses me if that is LLMs are so good at following instructions like this, why aren't they better with other structured data?
Like their inability to use the abstract in scientific papers as the required summary of its contents?
Like their inability to use the abstract in scientific papers as the required summary of its contents?
June 19, 2025 at 2:48 PM
What confuses me if that is LLMs are so good at following instructions like this, why aren't they better with other structured data?
Like their inability to use the abstract in scientific papers as the required summary of its contents?
Like their inability to use the abstract in scientific papers as the required summary of its contents?