MEMECAP: A Dataset for Captioning and Interpreting Memes
“We present MEMECAP, the first meme captioning dataset. MEMECAP is challenging for the existing VL models, as it requires recognizing and interpreting visual metaphors, and ignoring the literal visual elements. The experimental results using state-ofthe-art VL models indeed show that such models are still far from human performance. In particular, they tend to treat visual elements too literally and copy text from inside the meme.“
= Modern AIs still shockingly bad at understanding jokes, let alone creating them.
Though TBF: A shocking number of people also couldn’t properly explain a joke to save their lives.
Look at this, the paper’s own example of a good human explanation: “Meme poster finds it entertaining to read through long comment threads of arguments that happened in the past.” — Itself totally fails to explain the top essential property of any joke, surprise.
Worst mistake of jokes papers is to fail to consider that randomly-chosen human judges may themselves be objectively horrible at getting or explaining jokes.
Paper
Github
“We present MEMECAP, the first meme captioning dataset. MEMECAP is challenging for the existing VL models, as it requires recognizing and interpreting visual metaphors, and ignoring the literal visual elements. The experimental results using state-ofthe-art VL models indeed show that such models are still far from human performance. In particular, they tend to treat visual elements too literally and copy text from inside the meme.“
= Modern AIs still shockingly bad at understanding jokes, let alone creating them.
Though TBF: A shocking number of people also couldn’t properly explain a joke to save their lives.
Look at this, the paper’s own example of a good human explanation: “Meme poster finds it entertaining to read through long comment threads of arguments that happened in the past.” — Itself totally fails to explain the top essential property of any joke, surprise.
Worst mistake of jokes papers is to fail to consider that randomly-chosen human judges may themselves be objectively horrible at getting or explaining jokes.
Paper
Github
👏12❤4💯2🎉1👌1
Tide finally turning against the wordcel morons who repeat that there’s no way AIs could think because “it's just statistics bro”?
Daily reminder that “determines which word is statistically most likely to come next” — is an absolute lie.
This is not what modern RLHF’d LLMs do, at all.
Not every floating point number in the world is a “probability”.
Valuation in some valuation model, perhaps, but not a probability. Two very different things.
Let’s put this nonsense to bed.
Daily reminder that “determines which word is statistically most likely to come next” — is an absolute lie.
This is not what modern RLHF’d LLMs do, at all.
Not every floating point number in the world is a “probability”.
Valuation in some valuation model, perhaps, but not a probability. Two very different things.
Let’s put this nonsense to bed.
👍6👏4💯3❤1😐1
Where does the magic happen?
Some smart AI guys feel that it must occur at some lower level which they're unfamiliar with.
A single NAND is both extremely simple and achieves functional completeness — meaning it’s able to construct anything, including arbitrarily-intelligent thinking machines — but no, I assure you the magic is not happening at the NAND gate level.
So what is general intelligence, mathematically, logically?
Where does the magic happen?
I say, not just happening when the gates or weights are just sitting there, saved on disk -- but the magic is created when you dump massive amounts of resources into creating or running the AI, at training inference.
E.g. see blood flow to brains being far more predictive of intelligence in animals and humans than other measures like brain size.
Not just large, but obscenely large energy expendature that humans use just to think, so large that by itself this would kill many other animals from starvation.
I.e. Sufficiently obscene resource expenditure is indistinguishable from magic.
I.e., yet again, “The Bitter Lesson”, massive resource expenditure both makes the magic happen, and is the magic.
Functional completeness
Cerebral blood flow predicts multiple demand network activity and fluid intelligence across the adult lifespan.
Some smart AI guys feel that it must occur at some lower level which they're unfamiliar with.
A single NAND is both extremely simple and achieves functional completeness — meaning it’s able to construct anything, including arbitrarily-intelligent thinking machines — but no, I assure you the magic is not happening at the NAND gate level.
So what is general intelligence, mathematically, logically?
Where does the magic happen?
I say, not just happening when the gates or weights are just sitting there, saved on disk -- but the magic is created when you dump massive amounts of resources into creating or running the AI, at training inference.
E.g. see blood flow to brains being far more predictive of intelligence in animals and humans than other measures like brain size.
Not just large, but obscenely large energy expendature that humans use just to think, so large that by itself this would kill many other animals from starvation.
I.e. Sufficiently obscene resource expenditure is indistinguishable from magic.
I.e., yet again, “The Bitter Lesson”, massive resource expenditure both makes the magic happen, and is the magic.
Functional completeness
Cerebral blood flow predicts multiple demand network activity and fluid intelligence across the adult lifespan.
👏8❤2👍1
Massive Resources Are All you Need
Both for animals and machines.
Not about more complicated architecture.
Almost entirely about just dumping vastly more resources in, to let it do far more compute.
BuT tHaT’s NoT SuStAiNaBle!!
Really bro? Then you go ahead and be the first to constrict the bloodflow to your obscenely resource-hungry brain. Be the first to jump off of this “unsustainable” curve that your brain is sitting right at the top of.
Blood-Thirsty Brains Key To Evolution Of Human Intelligence
Bitter Lesson of AI Intelligence
Both for animals and machines.
Not about more complicated architecture.
Almost entirely about just dumping vastly more resources in, to let it do far more compute.
BuT tHaT’s NoT SuStAiNaBle!!
Really bro? Then you go ahead and be the first to constrict the bloodflow to your obscenely resource-hungry brain. Be the first to jump off of this “unsustainable” curve that your brain is sitting right at the top of.
Blood-Thirsty Brains Key To Evolution Of Human Intelligence
Bitter Lesson of AI Intelligence
👍8😁5💯4❤1👌1
Why the reverse Flynn Effect — Of IQ increasing for decades, but suddenly reversing ever since the 90’s?
Is it because we’re too addicted to tech which makes us lazy?
Immigration of dummies?
Climate change?
No.
Obesity, overwhelmingly.
Massively increased obesity in many countries → Obesity massively decreasing cerebral blood flow → which has an extremely strong negative effect on general intelligence → Massively decreased average intelligence.
Reverse flynn effect solved.
Brain needs power, obesity restricts it.
But hey, with human intelligence dropping so fast, this means we technically get to reach AGI that much sooner!
Who knew that the “singularity” was actually a reference to the size of yo momma on the day that AI finally surpasses mankind.
Tehnological Singularity
Yo Momma Singularity
Is it because we’re too addicted to tech which makes us lazy?
Immigration of dummies?
Climate change?
No.
Obesity, overwhelmingly.
Massively increased obesity in many countries → Obesity massively decreasing cerebral blood flow → which has an extremely strong negative effect on general intelligence → Massively decreased average intelligence.
Reverse flynn effect solved.
Brain needs power, obesity restricts it.
But hey, with human intelligence dropping so fast, this means we technically get to reach AGI that much sooner!
Who knew that the “singularity” was actually a reference to the size of yo momma on the day that AI finally surpasses mankind.
Tehnological Singularity
Yo Momma Singularity
👏22😁11👍10❤9🤯5❤🔥2🎉2😈2🔥1👌1💔1
LLMs to enable self-driving cars to consciously think and plan via internal monologue
“We use natural language to enhance the learning and explainability of our foundation driving models. In this blog, we introduce LINGO-1, an open-loop driving commentator that combines vision, language and action to enhance how we interpret, explain and train our foundation driving models.”
“We can also use language to probe models with questions about the driving scene to more intuitively understand what it comprehends. This capability can provide insights that could help us improve our driving models’ reasoning and decision-making capabilities. Equally exciting, VLAMs open up the possibility of interacting with driving models through dialogue, where users can ask autonomous vehicles what they are doing and why. This could significantly impact the public’s perception of this technology, building confidence and trust in its capabilities.”
“In addition to having a foundation driving model with broad capabilities, it is also eminently desirable for it to efficiently learn new tasks and quickly adapt to new domains and scenarios where we have small training samples. Here is where natural language could add value in supporting faster learning. For instance, we can imagine a scenario where a corrective driving action is accompanied by a natural language description of incorrect and correct behaviour in this situation. This extra supervision can enhance few-shot adaptations of the foundation model. With these ideas in mind, our Science team is exploring using natural language to build foundation models for end-to-end autonomous driving.”
LINGO-1: Exploring Natural Language for Autonomous Driving
“We use natural language to enhance the learning and explainability of our foundation driving models. In this blog, we introduce LINGO-1, an open-loop driving commentator that combines vision, language and action to enhance how we interpret, explain and train our foundation driving models.”
“We can also use language to probe models with questions about the driving scene to more intuitively understand what it comprehends. This capability can provide insights that could help us improve our driving models’ reasoning and decision-making capabilities. Equally exciting, VLAMs open up the possibility of interacting with driving models through dialogue, where users can ask autonomous vehicles what they are doing and why. This could significantly impact the public’s perception of this technology, building confidence and trust in its capabilities.”
“In addition to having a foundation driving model with broad capabilities, it is also eminently desirable for it to efficiently learn new tasks and quickly adapt to new domains and scenarios where we have small training samples. Here is where natural language could add value in supporting faster learning. For instance, we can imagine a scenario where a corrective driving action is accompanied by a natural language description of incorrect and correct behaviour in this situation. This extra supervision can enhance few-shot adaptations of the foundation model. With these ideas in mind, our Science team is exploring using natural language to build foundation models for end-to-end autonomous driving.”
LINGO-1: Exploring Natural Language for Autonomous Driving
🔥14👍4🤯4❤3💅2🎃1🤝1
AI Citations
How many sources dictating something to be true, does it take before a statement can accepted as truth?
How many sources dictating something to be true, does it take before a statement can accepted as truth?
Anonymous Poll
17%
1 good source dictating that something is true is acceptable
8%
2 good sources dictating that something is true is acceptable
13%
90% of sources, after considering some large number of sources
7%
99% of sources, after considering some large number of sources
29%
Infinite sources, never, no matter how many sources, no number of sources is ever enough
26%
Show results
❤8
OpenAI reminding everyone to switch from the less-censored text-davinci-003, to the heavily-censored, crippled gpt-3.5-turbo-instruct
Starting January 4, 2024, text-davinci-003 will no longer be available.
Announcement
Starting January 4, 2024, text-davinci-003 will no longer be available.
Announcement
😢14🤬8❤3
Is “validation” basically the same thing as “verification”?
Anonymous Poll
22%
YES SAME, validation & verification are essentially the same in contexts where it matters most.
44%
NO OPPOSITE, validation & verification are completely opposite in contexts where it matters most.
18%
Bro, I don’t even understand this question.
17%
Show results
👏6❤3👀3