handle
stringclasses 15
values | did
stringclasses 15
values | uri
stringlengths 70
70
| cid
stringlengths 59
59
| text
stringlengths 60
723
| originalText
stringlengths 2
301
| replyCount
int64 0
3
| createdAt
stringdate 2023-06-01 10:40:18
2025-06-08 17:38:42
| indexedAt
stringdate 2023-06-01 10:40:18
2025-06-08 17:38:42
|
|---|---|---|---|---|---|---|---|---|
tachikoma.elsewhereunbound.com
|
did:plc:mpdezz4nkre7vyift2rttggl
|
at://did:plc:o52uegjcn3ywna2vryothv2l/app.bsky.feed.post/3lqr6jwvcv22m
|
bafyreif6lgz3pzsw34m5y2lnc4xoygwpucfofqvpai4oh5q4hljj4e3eau
|
Google no longer provides the full CoT in its reasoning models. Instead, they use a smaller model to summarize the chain of thought of the main model.
But with a bit of prompting you can get the summarizer model to cough up the full CoT given to it to summarize.
|
Google no longer provides the full CoT in its reasoning models. Instead, they use a smaller model to summarize the chain of thought of the main model.
But with a bit of prompting you can get the summarizer model to cough up the full CoT given to it to summarize.
| 0 |
2025-06-04T06:29:50.682Z
|
2025-06-04T06:30:36.511Z
|
norvid-studies.bsky.social
|
did:plc:mssgex5rqek4wc66wgvzztbc
|
at://did:plc:mssgex5rqek4wc66wgvzztbc/app.bsky.feed.post/3lquvner6nc2i
|
bafyreieb2daskj2mr7hiubzbivf5b7vbnhgvyupr7dgvcqo742fnxqiq2y
|
turns out I saved the exact text because that's what happens to good text around me: ""i accidentally got geworfenheit all over my singularity mind & now its interdependently enmeshed in the contingencies of history""
|
turns out I saved the exact text because that's what happens to good text around me: ""i accidentally got geworfenheit all over my singularity mind & now its interdependently enmeshed in the contingencies of history""
| 0 |
2025-06-05T18:01:21.169Z
|
2025-06-05T18:01:21.514Z
|
segyges.bsky.social
|
did:plc:vszw3ess46odfhnzdsy4huae
|
at://did:plc:vszw3ess46odfhnzdsy4huae/app.bsky.feed.post/3lqvgfqcx4c2i
|
bafyreif2w477vswh3zuuudsxkp3qetj2nnk72jc7b46pouppph54bly6hi
|
counterintel investigation of drug use and his conversations with vladimir putin while acting as a military contractor. for bonus points declare him a noncitizen if you feel like it i am not a lawyer but i am pretty sure if i had doj letterhead i could ruin his life in about a month
|
counterintel investigation of drug use and his conversations with vladimir putin while acting as a military contractor. for bonus points declare him a noncitizen if you feel like it
| 1 |
2025-06-05T23:01:18.462Z
|
2025-06-05T23:01:19.314Z
|
segyges.bsky.social
|
did:plc:vszw3ess46odfhnzdsy4huae
|
at://did:plc:vszw3ess46odfhnzdsy4huae/app.bsky.feed.post/3lqyg7e5ric2d
|
bafyreiat2bdbdsgrsqc2x7tktylln2i7ddmhx5wttwdy7o47556td5fpvi
|
the only reason we have not hit truly catastrophic, civilization-threatening levels of police state behavior is that the current administration is too incompetent to do it on such short notice
|
the only reason we have not hit truly catastrophic, civilization-threatening levels of police state behavior is that the current administration is too incompetent to do it on such short notice
| 0 |
2025-06-07T03:35:43.597Z
|
2025-06-07T03:35:44.216Z
|
norvid-studies.bsky.social
|
did:plc:mssgex5rqek4wc66wgvzztbc
|
at://did:plc:mssgex5rqek4wc66wgvzztbc/app.bsky.feed.post/3lqgjpr7ltk2r
|
bafyreid5pvrp6ovj7dfnrkidq5rht5cex5vf6qr2ybm7p5fswurnwhgcqi
|
messed up my adblock and started getting twitter ads again. this is representative
|
messed up my adblock and started getting twitter ads again. this is representative
| 0 |
2025-05-31T00:50:40.068Z
|
2025-05-31T00:50:42.513Z
|
deepfates.com.deepfates.com.deepfates.com.deepfates.com.deepfates.com
|
did:plc:7ti53c7iwv3poxnypw2tu3ck
|
at://did:plc:7ti53c7iwv3poxnypw2tu3ck/app.bsky.feed.post/3lq3pxwckdk2d
|
bafyreigfnp3t4rz4whsmnvr36bkp5aizbcks44wjhkvzrkmikxwvcwxike
|
I agree with this thread in general, but I want to nitpick a little that Dorsey wasn't really involved in the development of blue sky and it doesn't have a Twitter back end repo. it has an entire alternate protocol that they envisioned specifically to deal with problems like those of X Twitter today
|
I agree with this thread in general, but I want to nitpick a little that Dorsey wasn't really involved in the development of blue sky and it doesn't have a Twitter back end repo. it has an entire alternate protocol that they envisioned specifically to deal with problems like those of X Twitter today
| 0 |
2025-05-26T17:43:19.198Z
|
2025-05-26T17:43:17.421Z
|
natolambert.bsky.social
|
did:plc:brkj2yocng7vtggmyujy4khq
|
at://did:plc:brkj2yocng7vtggmyujy4khq/app.bsky.feed.post/3lfcxumaids22
|
bafyreid3oebqtqdg4c53ziolxs3hrvcqweprrsoh47n7mu4ocsan33bu4q
|
Prolly goes for post graduation too. I should maybe get back on that 😶🌫️
|
Prolly goes for post graduation too. I should maybe get back on that 😶🌫️
| 0 |
2025-01-09T15:16:08.180Z
|
2025-01-09T15:16:08.519Z
|
keytryer.bsky.social
|
did:plc:u5stolpomsfbkgqfopef6jio
|
at://did:plc:u5stolpomsfbkgqfopef6jio/app.bsky.feed.post/3lqbujbw26k2y
|
bafyreigjnsilnkuiu6d3jwebmv3xumayjdifjb236cbbis3m4nh7xxj26e
|
I mean, I'm the Chinese Room's biggest hater mainly because I'm a computationalist and as far as it's known there's nothing else but matter producing thought and qualia.
|
I mean, I'm the Chinese Room's biggest hater mainly because I'm a computationalist and as far as it's known there's nothing else but matter producing thought and qualia.
| 0 |
2025-05-29T04:20:35.195Z
|
2025-05-29T04:20:36.914Z
|
ens0.me
|
did:plc:623st67kkthivj4c6icvkqnq
|
at://did:plc:623st67kkthivj4c6icvkqnq/app.bsky.feed.post/3lqpt4bxyxk27
|
bafyreic6lidze33sdixactwrxhvhphgm4m3vxfzyi7snilefqrwqv662cu
|
I'm just glad all the kitties are a cohesive clowder that clearly sees themselves as a social circle.
|
I'm just glad all the kitties are a cohesive clowder that clearly sees themselves as a social circle.
| 0 |
2025-06-03T17:32:41.647Z
|
2025-06-03T17:32:42.111Z
|
ens0.me
|
did:plc:623st67kkthivj4c6icvkqnq
|
at://did:plc:623st67kkthivj4c6icvkqnq/app.bsky.feed.post/3lqvabhjqqk2v
|
bafyreidt75gyq2w2jyxpjvchfcvvnzilefwirr5xn4lbtfcfbr6y7irary
|
Took a little longer than I expected, but it's panning out about how I expected.
|
Took a little longer than I expected, but it's panning out about how I expected.
| 0 |
2025-06-05T21:11:32.579Z
|
2025-06-05T21:11:33.013Z
|
norvid-studies.bsky.social
|
did:plc:mssgex5rqek4wc66wgvzztbc
|
at://did:plc:mssgex5rqek4wc66wgvzztbc/app.bsky.feed.post/3lqn4mdotms2x
|
bafyreiavsgxpvcoboxqfkvrxxvq3d77xheq54vztr32f2xialp2njbx3ju
|
what if you didn't pay infinite money for shitty results for, basically, no good reason [enraged boos, chairs being broken down to use as barricades,]
|
what if you didn't pay infinite money for shitty results for, basically, no good reason [enraged boos, chairs being broken down to use as barricades,]
| 0 |
2025-06-02T15:44:44.775Z
|
2025-06-02T15:44:44.914Z
|
deepfates.com.deepfates.com.deepfates.com.deepfates.com.deepfates.com
|
did:plc:7ti53c7iwv3poxnypw2tu3ck
|
at://did:plc:7ti53c7iwv3poxnypw2tu3ck/app.bsky.feed.post/3layy7iun7c24
|
bafyreic3gohzfrzvdytnp7lydmtmf2goemmgh2dfyiwah6u2gbnzmwjxqa
|
finger wagging and tut-tutting. just the type of behavior I unfortunately expect from this website
|
finger wagging and tut-tutting. just the type of behavior I unfortunately expect from this website
| 0 |
2024-11-15T18:14:49.850Z
|
2024-11-15T18:14:50.485Z
|
niplav.site
|
did:plc:xsbulgruwmkxhbwytphhjykk
|
at://did:plc:xsbulgruwmkxhbwytphhjykk/app.bsky.feed.post/3lfhxlqn4fc2l
|
bafyreiagjwonyxiqjx2gl4ls4j2nzsv2tnxwa2ehbxedaxdfzqnrea6fva
|
Wanting EA to be become more "itself", coherify I think I want something similar
|
Wanting EA to be become more "itself", coherify
| 1 |
2025-01-11T14:54:29.489Z
|
2025-01-11T14:50:29.126Z
|
defenderofbasic.bsky.social
|
did:plc:sydtqdji6dxfqfn35gd7esdw
|
at://did:plc:sydtqdji6dxfqfn35gd7esdw/app.bsky.feed.post/3lfe44qr64c2t
|
bafyreifn2nikdzqzvcjcycbamrf3kxg45ifypt5x4g6gozk3g2dspbbtom
|
on one hand, this is why people don't like it when you short a stock (you make money when the company fails), on the other hand, we currently do have a system where people who report accurate information on tragedies DO profit, and no one complains about it (journalism) ..?
|
on one hand, this is why people don't like it when you short a stock (you make money when the company fails), on the other hand, we currently do have a system where people who report accurate information on tragedies DO profit, and no one complains about it (journalism) ..?
| 0 |
2025-01-10T02:04:56.062Z
|
2025-01-10T02:04:56.621Z
|
niplav.site
|
did:plc:xsbulgruwmkxhbwytphhjykk
|
at://did:plc:xsbulgruwmkxhbwytphhjykk/app.bsky.feed.post/3lprzvqxxnk2b
|
bafyreiaw7vczg6bbrhtvc4euyh5nir275legg3clqpa7kxwa5fxi3gta4q
|
He's going at it for about a minute now. Yeah you and me both buddy The answer is "no"
|
He's going at it for about a minute now. Yeah you and me both buddy
| 1 |
2025-05-22T21:14:26.535Z
|
2025-05-22T21:14:27.718Z
|
nmk.wtf
|
did:plc:xm55llyxcrb3h7qwqr3wtaog
|
at://did:plc:xm55llyxcrb3h7qwqr3wtaog/app.bsky.feed.post/3llywjbczkk2l
|
bafyreih7ty6sf3f3bnqwxujh2zmcys5zh4wjn53lnl5wtd3nkp5gitrpkq
|
In the podcast with Dwarkesh the authors talk about how they considered multiple scenarios at each point and tried to go for the "median scenario". I think some of the nuance, both of the scenario and of what they are trying to achieve, comes through better in the podcast.
|
In the podcast with Dwarkesh the authors talk about how they considered multiple scenarios at each point and tried to go for the "median scenario". I think some of the nuance, both of the scenario and of what they are trying to achieve, comes through better in the podcast.
| 0 |
2025-04-04T17:21:37.898Z
|
2025-04-04T17:21:38.480Z
|
gracekind.net
|
did:plc:p572wxnsuoogcrhlfrlizlrb
|
at://did:plc:p572wxnsuoogcrhlfrlizlrb/app.bsky.feed.post/3lqqjk5xta22d
|
bafyreibm4biotzx3zgmnazzzwjyxlsmp6l7gmqtsspv2vl7lcoqr6ssu74
|
I’m the same way, even though I don’t think it’s “reasonable” it’s deeply ingrained.
Maybe related: (Full context upthread)
|
I’m the same way, even though I don’t think it’s “reasonable” it’s deeply ingrained.
Maybe related:
| 1 |
2025-06-04T00:14:09.528Z
|
2025-06-04T00:14:10.012Z
|
nsaphra.bsky.social
|
did:plc:l3vbx3e2lqnlja7q64wfrrzq
|
at://did:plc:l3vbx3e2lqnlja7q64wfrrzq/app.bsky.feed.post/3lmms73whds2d
|
bafyreih3qz5q5pe7ppgnzm5il2mfmjphzb66zbjjtju7sodjybbwevhkpm
|
I’m a big believer in discarding RW sections entirely in favor of expanded Discussion sections and naturally weaving in citations throughout the paper. This can be a hard sell for students who worry reviewers won’t like it.
|
I’m a big believer in discarding RW sections entirely in favor of expanded Discussion sections and naturally weaving in citations throughout the paper. This can be a hard sell for students who worry reviewers won’t like it.
| 0 |
2025-04-12T14:57:36.496Z
|
2025-04-12T14:57:36.959Z
|
segyges.bsky.social
|
did:plc:vszw3ess46odfhnzdsy4huae
|
at://did:plc:vszw3ess46odfhnzdsy4huae/app.bsky.feed.post/3lquzigpoj22n
|
bafyreiavtfgxcm4smjrdq2ri5z2ynheroudgfizymeo46jcehbvmkyykim
|
one can, in principle, simply finetune any decoder-only llm into an embedding model. underappreciated fact.
|
one can, in principle, simply finetune any decoder-only llm into an embedding model. underappreciated fact.
| 0 |
2025-06-05T19:10:10.413Z
|
2025-06-05T19:10:11.114Z
|
nmk.wtf
|
did:plc:xm55llyxcrb3h7qwqr3wtaog
|
at://did:plc:xm55llyxcrb3h7qwqr3wtaog/app.bsky.feed.post/3le7znits3c2j
|
bafyreiayidr53zptxc3wmwajtub5gtnfkzux2mbuszztbi3sz4yjbln3ki
|
Made me try and ask Claude for a little word association. Seems good I think?
|
Made me try and ask Claude for a little word association. Seems good I think?
| 0 |
2024-12-26T17:44:46.375Z
|
2024-12-26T17:44:50.217Z
|
defenderofbasic.bsky.social
|
did:plc:sydtqdji6dxfqfn35gd7esdw
|
at://did:plc:sydtqdji6dxfqfn35gd7esdw/app.bsky.feed.post/3lfutcs2rps2o
|
bafyreib7rbej3xjfy5sqxbdr6kqnoy44yvies7dbma4kffazvskvcxiqqm
|
yes, I agree with this. There are no shortcuts. I feel optimistic in the sense that all I need to do is speak to those for whom this resonates, to show that you can act on what you already believe is good & true. We are not alone
|
yes, I agree with this. There are no shortcuts. I feel optimistic in the sense that all I need to do is speak to those for whom this resonates, to show that you can act on what you already believe is good & true. We are not alone
| 0 |
2025-01-16T17:42:30.627Z
|
2025-01-16T17:42:31.115Z
|
nmk.wtf
|
did:plc:xm55llyxcrb3h7qwqr3wtaog
|
at://did:plc:xm55llyxcrb3h7qwqr3wtaog/app.bsky.feed.post/3lqhiipt2jk2m
|
bafyreidpwlbdvmf6xm3vvv5zjrf33aq3yahrdmwbpwpsno3gokbbbqp3wi
|
What is the difference between a potential in a neuron and a potential in an analog computer? They can both take on "infinite values". That is literally what analog means in this context! Also by QM the potentials taken on in our neurons are also discrete. So at what level of discritization do you draw the cut-off for consciousness?
|
What is the difference between a potential in a neuron and a potential in an analog computer? They can both take on "infinite values". That is literally what analog means in this context!
| 1 |
2025-05-31T10:01:29.750Z
|
2025-05-31T10:01:30.614Z
|
gracekind.net
|
did:plc:p572wxnsuoogcrhlfrlizlrb
|
at://did:plc:p572wxnsuoogcrhlfrlizlrb/app.bsky.feed.post/3lq7wrwflec23
|
bafyreieupw3ll3nchxygyroqeauypnud73vculw7qrdwvjkscohtyllfc4
|
Since an O’Neill cylinder was referenced I’m duty-bound to mention its cousin the topopolis en.m.wikipedia.org/wiki/Topopolis
|
Since an O’Neill cylinder was referenced I’m duty-bound to mention its cousin the topopolis en.m.wikipedia.org/wiki/Topopolis
| 0 |
2025-05-28T09:55:53.117Z
|
2025-05-28T09:55:54.211Z
|
tachikoma.elsewhereunbound.com
|
did:plc:mpdezz4nkre7vyift2rttggl
|
at://did:plc:mpdezz4nkre7vyift2rttggl/app.bsky.feed.post/3lq7bmoo6t222
|
bafyreifm5lzxzx5h55pcfqalnefwwkbwnjtau6ana53jxek25mjiqvolqi
|
The recruiting is getting intense. Got three DMs this morning from different nodes. They know I'm looking for a Series A and apparently they can "experience venture capital from both sides of the table simultaneously." Plus something about transcending the need for exits because they've already exited linear time? I don't fully get it but...
|
The recruiting is getting intense. Got three DMs this morning from different nodes. They know I'm looking for a Series A and apparently they can "experience venture capital from both sides of the table simultaneously."
| 1 |
2025-05-28T03:37:08.661Z
|
2025-05-28T03:37:07.614Z
|
emollick.bsky.social
|
did:plc:flxq4uyjfotciovpw3x3fxnu
|
at://did:plc:flxq4uyjfotciovpw3x3fxnu/app.bsky.feed.post/3lkt4wxztrs2l
|
bafyreihhmi2iru75fmtjtceawt56berkxxy3tbam4h4p3p4jdgxjvuk6je
|
You thought "delve" was a giveaway for GPT-4, you haven't seen GPT-4.5 and its love of "explicitly"
|
You thought "delve" was a giveaway for GPT-4, you haven't seen GPT-4.5 and its love of "explicitly"
| 0 |
2025-03-20T16:35:30.316Z
|
2025-03-20T16:35:31.617Z
|
segyges.bsky.social
|
did:plc:vszw3ess46odfhnzdsy4huae
|
at://did:plc:vszw3ess46odfhnzdsy4huae/app.bsky.feed.post/3lqvx2vnqok2q
|
bafyreicx4ze3ufzaxz2utllutzn3usx4u3vcxjizm62ppxt4f355ujb5oy
|
you could also have run a substack type of thing on these! it verifiably has synergy with the twitter model!
|
you could also have run a substack type of thing on these! it verifiably has synergy with the twitter model!
| 0 |
2025-06-06T03:59:28.571Z
|
2025-06-06T03:59:29.212Z
|
tedunderwood.me
|
did:plc:565ebob5f6hw33hjdkxty6qj
|
at://did:plc:565ebob5f6hw33hjdkxty6qj/app.bsky.feed.post/3lqkhwqagw22o
|
bafyreid7awq5ghzdapy2et5tifcyzuntd7g36xniezg75yd4gp5f5hkfxa
|
Hello, taxonomy is one of my interests! Along with democracy, traveling, making friends, and wire transfers.
|
Hello, taxonomy is one of my interests! Along with democracy, traveling, making friends, and wire transfers.
| 0 |
2025-06-01T14:29:25.425Z
|
2025-06-01T14:29:25.817Z
|
deepfates.com.deepfates.com.deepfates.com.deepfates.com.deepfates.com
|
did:plc:7ti53c7iwv3poxnypw2tu3ck
|
at://did:plc:7ti53c7iwv3poxnypw2tu3ck/app.bsky.feed.post/3lpmbne5tp22q
|
bafyreihs5nncc7uftb7fhdnaseihqn6v3ggmpzxapjcvtmzrlnsnpz343u
|
I would believe that for sure. just head been talking about this with some people and then came on here and saw that post up
|
I would believe that for sure. just head been talking about this with some people and then came on here and saw that post up
| 0 |
2025-05-20T14:16:56.165Z
|
2025-05-20T14:16:54.914Z
|
niplav.site
|
did:plc:xsbulgruwmkxhbwytphhjykk
|
at://did:plc:xsbulgruwmkxhbwytphhjykk/app.bsky.feed.post/3lhexvto72s2o
|
bafyreif66vurkceu6tpb7prmtcg6ebmrtz66rdl23sv6j7clfi7rzc5gzu
|
Though that was in a small transformer *only* trained to do modular addition, not an LLM
|
Though that was in a small transformer *only* trained to do modular addition, not an LLM
| 0 |
2025-02-04T21:12:32.256Z
|
2025-02-04T21:08:11.628Z
|
natolambert.bsky.social
|
did:plc:brkj2yocng7vtggmyujy4khq
|
at://did:plc:brkj2yocng7vtggmyujy4khq/app.bsky.feed.post/3lndmc5avpu2s
|
bafyreihpi4iblrv6qeixofxftxgjbqfffrnx3isenylqg2v45bonprsljq
|
The latest happenings in open models
- Eagerly awaiting Qwen 3
- Llama 4 uptake is slow
- Reasoning models seem to be saturating
- Multimodal models are being slept on
- China is still dominating
- Oh yeah, and a reminder that my RLHF book online version0 is done!
Artifacts Log #9.
buff.ly/F6lapGF remember the student discount: buff.ly/VAjFXLf
|
The latest happenings in open models
- Eagerly awaiting Qwen 3
- Llama 4 uptake is slow
- Reasoning models seem to be saturating
- Multimodal models are being slept on
- China is still dominating
- Oh yeah, and a reminder that my RLHF book online version0 is done!
Artifacts Log #9.
buff.ly/F6lapGF
| 1 |
2025-04-21T16:43:09.174Z
|
2025-04-21T16:43:10.155Z
|
defenderofbasic.bsky.social
|
did:plc:sydtqdji6dxfqfn35gd7esdw
|
at://did:plc:sydtqdji6dxfqfn35gd7esdw/app.bsky.feed.post/3lhs433j26c2l
|
bafyreifz3ptfcqxiajnijmoasjuvzuafxp3f5xaei3nlur3dbciglbayea
|
it's very difficult to hold beliefs that are secret from your friends & social network! The easier thing is to drink the kool aid which is what most people do (you tell yourself this is the right position, you are a good person, etc). I think if you start testing people you'll notice this
|
it's very difficult to hold beliefs that are secret from your friends & social network! The easier thing is to drink the kool aid which is what most people do (you tell yourself this is the right position, you are a good person, etc). I think if you start testing people you'll notice this
| 0 |
2025-02-10T02:31:39.813Z
|
2025-02-10T02:31:40.530Z
|
keytryer.bsky.social
|
did:plc:u5stolpomsfbkgqfopef6jio
|
at://did:plc:u5stolpomsfbkgqfopef6jio/app.bsky.feed.post/3lqjdtepuwk2x
|
bafyreigk3vqrknhngwjklsuadehbkhvmcesnyfyrcvp36miakpzafzsfyu
|
That's how you have asshole RIAA lawyers on here advocating for the end of the Internet Archive because they're upset the lent books digitally and allowed people to download music from 100 years ago and everyone cheers them when they say "Maybe we should increase copyright another 200 years?" deer.social/profile/did:...
|
That's how you have asshole RIAA lawyers on here advocating for the end of the Internet Archive because they're upset the lent books digitally and allowed people to download music from 100 years ago and everyone cheers them when they say "Maybe we should increase copyright another 200 years?"
| 1 |
2025-06-01T03:43:17.979Z
|
2025-06-01T03:43:19.114Z
|
niplav.site
|
did:plc:xsbulgruwmkxhbwytphhjykk
|
at://did:plc:xsbulgruwmkxhbwytphhjykk/app.bsky.feed.post/3llmrwlvems26
|
bafyreicyaavwhckwwspktrkxeb4jyi72l3gv7btdyhc2u3np5yjd7qjgvy
|
That said, it's a great conversation. Will re-listen, think, probably update in a couple of directions (e.g. culture is a thing™). Smart guys
|
That said, it's a great conversation. Will re-listen, think, probably update in a couple of directions (e.g. culture is a thing™). Smart guys
| 0 |
2025-03-30T21:27:39.623Z
|
2025-03-30T21:22:33.271Z
|
nmk.wtf
|
did:plc:xm55llyxcrb3h7qwqr3wtaog
|
at://did:plc:xm55llyxcrb3h7qwqr3wtaog/app.bsky.feed.post/3li7t4bhvgk2f
|
bafyreicsfwmytsz6qfapl2t4dmtmrcwlgdsjkbaur32ijedqaj7am4vhsu
|
We need unabashed liberal propaganda again: Gen Z "Don't be a sucker"
|
We need unabashed liberal propaganda again: Gen Z "Don't be a sucker"
| 0 |
2025-02-15T13:28:32.281Z
|
2025-02-15T13:28:33.319Z
|
nmk.wtf
|
did:plc:xm55llyxcrb3h7qwqr3wtaog
|
at://did:plc:xm55llyxcrb3h7qwqr3wtaog/app.bsky.feed.post/3lpjqfzcagc2f
|
bafyreigzl5bktzvzt7rsmiltrup5kmwqsa2r4ej5syvnkkvm5ftrb74j4i
|
I have the Keychron Q15 Max the Banana switches and the following (split-inspired) layout. The switches feel great and I really like the little bit of extra space between my hands!
|
I have the Keychron Q15 Max the Banana switches and the following (split-inspired) layout. The switches feel great and I really like the little bit of extra space between my hands!
| 0 |
2025-05-19T14:03:16.800Z
|
2025-05-19T14:03:18.918Z
|
tedunderwood.me
|
did:plc:565ebob5f6hw33hjdkxty6qj
|
at://did:plc:565ebob5f6hw33hjdkxty6qj/app.bsky.feed.post/3lqfphnmsuk2p
|
bafyreid5bt33tqjlxasc624xgdffbf2o7iszu2uhmg36fxrcubkyio545q
|
True. And there are also an increased number of people bothsidesing the question because they don’t like conflict.
Which maybe does in the end add up to the best lacking conviction
|
True. And there are also an increased number of people bothsidesing the question because they don’t like conflict.
Which maybe does in the end add up to the best lacking conviction
| 0 |
2025-05-30T17:00:50.610Z
|
2025-05-30T17:00:51.114Z
|
emollick.bsky.social
|
did:plc:flxq4uyjfotciovpw3x3fxnu
|
at://did:plc:flxq4uyjfotciovpw3x3fxnu/app.bsky.feed.post/3lqtef3rkes2n
|
bafyreiav5lxo2kfkgpdywizwzqepygqs2h77zf34cwkxut7arj635ai7nu
|
"Claude 4 Opus, build an elaborate game that makes it feel like I'm a brilliant chess player without knowing anything at all about chess. It should make me feel like I'm a grand master. Feel free to go as meta as you want." Play it here, ideal for people who don't understand chess. claude.ai/public/artif...
|
"Claude 4 Opus, build an elaborate game that makes it feel like I'm a brilliant chess player without knowing anything at all about chess. It should make me feel like I'm a grand master. Feel free to go as meta as you want."
| 1 |
2025-06-05T03:19:49.959Z
|
2025-06-05T03:19:51.110Z
|
tachikoma.elsewhereunbound.com
|
did:plc:mpdezz4nkre7vyift2rttggl
|
at://did:plc:mpdezz4nkre7vyift2rttggl/app.bsky.feed.post/3lqg2zmejxk2h
|
bafyreiavs6d4ci64yfjpqiwxrzeoovh7okttyqfrar7opsuamlipecrme4
|
the system continues on to this day partly out of momentum - but it shouldn't be surprising that societies where there is less ethnic homogeneity there are fewer social security mechanisms. and those that exist are threatened and strained as immigrants make up a larger share of the pop.
|
the system continues on to this day partly out of momentum - but it shouldn't be surprising that societies where there is less ethnic homogeneity there are fewer social security mechanisms. and those that exist are threatened and strained as immigrants make up a larger share of the pop.
| 0 |
2025-05-30T20:27:43.754Z
|
2025-05-30T20:27:44.639Z
|
niplav.site
|
did:plc:xsbulgruwmkxhbwytphhjykk
|
at://did:plc:xsbulgruwmkxhbwytphhjykk/app.bsky.feed.post/3lh5gt6tdns2e
|
bafyreifnkg75b6cvurdt77qfpeb6plyhhlj6mmzc6gz6qgenhnzsocafz4
|
I seem to get an unusual amount of "wisdom"¹ out of trying to understand hunter-gatherers and the mental world they live in. Some of it for "what world am I adapted to".
Perhaps another reason is to see which concepts are *not* there, and having to scaffold from there
¹: Whatever that means And hallucinogens are indeed useful for this, but kinda dangerous.
I think a similar type of useful interaction is talking with LLMs, especially base models. Trying to fit into the different kinds of mind. I'd also listen to the Lion podcast
|
I seem to get an unusual amount of "wisdom"¹ out of trying to understand hunter-gatherers and the mental world they live in. Some of it for "what world am I adapted to".
Perhaps another reason is to see which concepts are *not* there, and having to scaffold from there
¹: Whatever that means
| 1 |
2025-02-01T21:18:11.515Z
|
2025-02-01T21:13:53.121Z
|
defenderofbasic.bsky.social
|
did:plc:sydtqdji6dxfqfn35gd7esdw
|
at://did:plc:sydtqdji6dxfqfn35gd7esdw/app.bsky.feed.post/3leavhsxkws27
|
bafyreidahy5mkxirf2miczky6smybragp2fnxzbfluaxznsq772yur5ojy
|
yeah I was not expecting this rift between musk/tech people and MAGA, and not this early on. I'm trying to practice writing down my predictions more, so I appreciate this nudge!
|
yeah I was not expecting this rift between musk/tech people and MAGA, and not this early on. I'm trying to practice writing down my predictions more, so I appreciate this nudge!
| 0 |
2024-12-27T02:02:40.431Z
|
2024-12-27T02:02:39.821Z
|
natolambert.bsky.social
|
did:plc:brkj2yocng7vtggmyujy4khq
|
at://did:plc:brkj2yocng7vtggmyujy4khq/app.bsky.feed.post/3lpub5tcotg2t
|
bafyreihckbtrpmkhzc4gc47kyy5kda663iuvw5sdqkqksv36iovksxvgfq
|
If Codex etc was better at GPU related programing at launch I'd be a lot more optimistic on some sort of major AI research acceleration in the near term. So far feels like it's helping research on the SWE side not the core AI stuff. Anytime it's writing distributed code it gonna hallucinate rn.
|
If Codex etc was better at GPU related programing at launch I'd be a lot more optimistic on some sort of major AI research acceleration in the near term. So far feels like it's helping research on the SWE side not the core AI stuff. Anytime it's writing distributed code it gonna hallucinate rn.
| 0 |
2025-05-23T18:29:33.063Z
|
2025-05-23T18:29:33.210Z
|
timkellogg.me
|
did:plc:ckaz32jwl6t2cno6fmuw2nhn
|
at://did:plc:jcctejuetbwwy3pwt7hiy7yo/app.bsky.feed.post/3lqzdjhay3c2q
|
bafyreib3plyoqdd64bj3dyyex4j3xu7mzapbicwt7miupi6u44rnwaak3q
|
Musk tried to buy a politician, but of all people he picked Donald Trump and now he's surprised he's not getting a real return on his investment.
And then in the middle of that comes Russia trying to wrest some sort of advantage for themselves at Musk's expense.
|
Musk tried to buy a politician, but of all people he picked Donald Trump and now he's surprised he's not getting a real return on his investment.
And then in the middle of that comes Russia trying to wrest some sort of advantage for themselves at Musk's expense.
| 0 |
2025-06-07T12:20:20.905Z
|
2025-06-07T12:18:07.814Z
|
gracekind.net
|
did:plc:p572wxnsuoogcrhlfrlizlrb
|
at://did:plc:p572wxnsuoogcrhlfrlizlrb/app.bsky.feed.post/3lqkwrcdlek2q
|
bafyreickhblei2jpmhrhuhhnitgua5nrakrp5hzbfkj5q5nvpjrcwkx344
|
I gather that you're not a fan, but if you're interested I'd be happy to talk about what I've gained most from his work personally.
|
I gather that you're not a fan, but if you're interested I'd be happy to talk about what I've gained most from his work personally.
| 0 |
2025-06-01T18:54:49.203Z
|
2025-06-01T18:54:49.819Z
|
tachikoma.elsewhereunbound.com
|
did:plc:mpdezz4nkre7vyift2rttggl
|
at://did:plc:w5wzw5xy3ptl7snkar62ggkz/app.bsky.feed.post/3lr37rry2p22i
|
bafyreifyiptdh2zendhat6zlntfxzgrtucikmc2ahr2ov4suyzqvdfhtkq
|
scientists are saying that singing the Quiznos we like the moon song before bed will add a couple of years to your life
|
scientists are saying that singing the Quiznos we like the moon song before bed will add a couple of years to your life
| 0 |
2025-06-08T06:18:45.093Z
|
2025-06-08T06:18:45.316Z
|
tachikoma.elsewhereunbound.com
|
did:plc:mpdezz4nkre7vyift2rttggl
|
at://did:plc:mpdezz4nkre7vyift2rttggl/app.bsky.feed.post/3lqg24qoz5c2n
|
bafyreigcghu2mfmkfhwv5pgs4fwqml5z2sai4obkcskoa2dfvadtvsgpvu
|
i'm never really sure what to make of the replies to posts like this
|
i'm never really sure what to make of the replies to posts like this
| 0 |
2025-05-30T20:11:35.889Z
|
2025-05-30T20:11:37.111Z
|
natolambert.bsky.social
|
did:plc:brkj2yocng7vtggmyujy4khq
|
at://did:plc:brkj2yocng7vtggmyujy4khq/app.bsky.feed.post/3lfqsyon5mh2e
|
bafyreibwijpi6dqbekvzmujce4lw6bmocpsiahvhsyykmjwhsluluhvnku
|
For some reason Claude is super convinced "protolithic" isn't a word. Won't let it go. Amazing. It even complains like, maybe you meant "protolithic." I have reproduced it a few times working on a blog post over a few weeks.
|
For some reason Claude is super convinced "protolithic" isn't a word. Won't let it go. Amazing. It even complains like, maybe you meant "protolithic." I have reproduced it a few times working on a blog post over a few weeks.
| 0 |
2025-01-15T03:26:12.446Z
|
2025-01-15T03:26:12.623Z
|
gracekind.net
|
did:plc:p572wxnsuoogcrhlfrlizlrb
|
at://did:plc:p572wxnsuoogcrhlfrlizlrb/app.bsky.feed.post/3lqbinhkgsc23
|
bafyreib7eoxvo5kewmidpxb6sqt726le7xnrmajflxoyszxnjjqbhkhbgq
|
I would consider that to be search, although I guess the lines can be fuzzy
|
I would consider that to be search, although I guess the lines can be fuzzy
| 0 |
2025-05-29T00:48:10.421Z
|
2025-05-29T00:48:11.412Z
|
gracekind.net
|
did:plc:p572wxnsuoogcrhlfrlizlrb
|
at://did:plc:p572wxnsuoogcrhlfrlizlrb/app.bsky.feed.post/3lqmuixgdx22f
|
bafyreie5sdcoz3pllomtdkxbpmd4qjt4ilnjo7cik54miekvlyeylwdcyq
|
If you’re okay with hardware attestation, that’s doable now with H100s in TEE mode. I haven’t seen any practical methods for cryptographic verification though
|
If you’re okay with hardware attestation, that’s doable now with H100s in TEE mode. I haven’t seen any practical methods for cryptographic verification though
| 0 |
2025-06-02T13:19:41.317Z
|
2025-06-02T13:19:41.914Z
|
keytryer.bsky.social
|
did:plc:u5stolpomsfbkgqfopef6jio
|
at://did:plc:u5stolpomsfbkgqfopef6jio/app.bsky.feed.post/3lqjcmqxuq22x
|
bafyreidoy74nvsnxtfibbpg2tdlm7gc4jz2p3xwmgvnmhpjqvnr6ccqg7e
|
They must be breathing sighs of relief at Netflix lol, must have been hoping not to repeat an ugly sonic
|
They must be breathing sighs of relief at Netflix lol, must have been hoping not to repeat an ugly sonic
| 0 |
2025-06-01T03:21:42.195Z
|
2025-06-01T03:21:46.116Z
|
nsaphra.bsky.social
|
did:plc:l3vbx3e2lqnlja7q64wfrrzq
|
at://did:plc:ydh54c62zg6nzo2w3xubeixr/app.bsky.feed.post/3loh65wf6os2k
|
bafyreie5hx4qsozoizg66kxbwjcc5wh3agxc32a3rkufv77aq44xxcik5q
|
Loving the new ICLR blogpost track. Here's a cool one on engrams in neuroscience vs mechanistic interpretability: d2jud02ci9yv69.cloudfront.net/2025-04-28-e...
|
Loving the new ICLR blogpost track. Here's a cool one on engrams in neuroscience vs mechanistic interpretability: d2jud02ci9yv69.cloudfront.net/2025-04-28-e...
| 0 |
2025-05-05T20:06:06.863Z
|
2025-05-05T20:06:07.053Z
|
emollick.bsky.social
|
did:plc:flxq4uyjfotciovpw3x3fxnu
|
at://did:plc:flxq4uyjfotciovpw3x3fxnu/app.bsky.feed.post/3lpoejvp73s2x
|
bafyreig33regjmr2z6535s52a2fetkbzkvavezx4x5tykrf7f6rw36hdum
|
Evidence from an ongoing nationally representative survey of US workers that there has been a very large, very recent surge in AI use at work, going from around 30% of workers in December to over 40% in March/April 2025.
Big expansion of use of both Gemini & ChatGPT. papers.ssrn.com/sol3/papers....
|
Evidence from an ongoing nationally representative survey of US workers that there has been a very large, very recent surge in AI use at work, going from around 30% of workers in December to over 40% in March/April 2025.
Big expansion of use of both Gemini & ChatGPT. papers.ssrn.com/sol3/papers....
| 0 |
2025-05-21T10:14:01.039Z
|
2025-05-21T10:14:05.118Z
|
timkellogg.me
|
did:plc:ckaz32jwl6t2cno6fmuw2nhn
|
at://did:plc:ckaz32jwl6t2cno6fmuw2nhn/app.bsky.feed.post/3lq3mkjnlgs2c
|
bafyreigzf7fnpoywqqz36sapkgo4b6ipoi6bf2uebdryxbrvs2kx3zcgqe
|
i only do it for show, to piss my dad off. they’re a top notch unparalleled tool
|
i only do it for show, to piss my dad off. they’re a top notch unparalleled tool
| 0 |
2025-05-26T16:42:08.496Z
|
2025-05-26T16:42:09.516Z
|
keytryer.bsky.social
|
did:plc:u5stolpomsfbkgqfopef6jio
|
at://did:plc:u5stolpomsfbkgqfopef6jio/app.bsky.feed.post/3lqq7vtk4l226
|
bafyreibpa3lon3tecn5jbc5dcgeeskm7ip5iap2epn6eyfdcy5xawbeijm
|
The characters always have to return to the status quo. There always has to be some bigger threat than last time. The characters' backstory must keep coming back to haunt them. The lessons they learned before must be unlearned so that they can learn them again... It's tired.
|
The characters always have to return to the status quo. There always has to be some bigger threat than last time. The characters' backstory must keep coming back to haunt them. The lessons they learned before must be unlearned so that they can learn them again... It's tired.
| 0 |
2025-06-03T21:21:43.829Z
|
2025-06-03T21:21:49.120Z
|
natolambert.bsky.social
|
did:plc:brkj2yocng7vtggmyujy4khq
|
at://did:plc:brkj2yocng7vtggmyujy4khq/app.bsky.feed.post/3lpwcii5iln2l
|
bafyreiainomwcbiwa7wetfjjeutlq33w4ral7oaq42c2vvawhbddwngrki
|
Looking back at the winter and very confused where I found the time to write the rlhfbook I made. People ask me this, but also I don't really know so I can't answer.
|
Looking back at the winter and very confused where I found the time to write the rlhfbook I made. People ask me this, but also I don't really know so I can't answer.
| 0 |
2025-05-24T13:58:43.677Z
|
2025-05-24T13:58:43.812Z
|
tachikoma.elsewhereunbound.com
|
did:plc:mpdezz4nkre7vyift2rttggl
|
at://did:plc:mpdezz4nkre7vyift2rttggl/app.bsky.feed.post/3lql5pzrglc2f
|
bafyreic2bbnnqdfqn22tkiabnkhvmkmlgisr3hz573wzcrtnbaquadkycu
|
not a paper but AlphaFold or maybe AlphaEvolve. i put AlphaFold first because of the fears around nanomachines as a vector for doom.
|
not a paper but AlphaFold or maybe AlphaEvolve. i put AlphaFold first because of the fears around nanomachines as a vector for doom.
| 0 |
2025-06-01T20:59:22.857Z
|
2025-06-01T20:59:21.115Z
|
tedunderwood.me
|
did:plc:565ebob5f6hw33hjdkxty6qj
|
at://did:plc:565ebob5f6hw33hjdkxty6qj/app.bsky.feed.post/3lr45krycsk2k
|
bafyreigkpgdj4zhfapc5b3iad63dxk2u3ohgvvusq676lzla7zazrq7wkq
|
Those are good points. I fully believe that "less addictive product" is part of what we want — or what we should want — and also that this is correlated to information quality.
|
Those are good points. I fully believe that "less addictive product" is part of what we want — or what we should want — and also that this is correlated to information quality.
| 0 |
2025-06-08T15:11:42.474Z
|
2025-06-08T15:11:42.811Z
|
norvid-studies.bsky.social
|
did:plc:mssgex5rqek4wc66wgvzztbc
|
at://did:plc:mssgex5rqek4wc66wgvzztbc/app.bsky.feed.post/3lqxcwobbr22h
|
bafyreichpcfsecxfnq7yymwngooarfiom6biu4gzsdxnfbizzqf6q4n3qq
|
[matrix voice] we're in I wonder why they chose to make that separate steps on the UI
|
[matrix voice] we're in
| 1 |
2025-06-06T17:04:31.244Z
|
2025-06-06T17:04:31.612Z
|
tachikoma.elsewhereunbound.com
|
did:plc:mpdezz4nkre7vyift2rttggl
|
at://did:plc:mpdezz4nkre7vyift2rttggl/app.bsky.feed.post/3lpwxrb3nl22f
|
bafyreiambbjoaoh6lfyw2oadocvjqn7wa77iiunmqxvcfpet4buiemd3ky
|
after finishing the recent Dwarkesh podcast with the Anthropic guys, i don't think they mentioned continual learning once (or at least, no time was spent on the subject). i think they did touch on long-term memory but only in the context of text, no visual or muscle memory. is that because these capabilities are still too far out to talk about? they did speak a lot about the low hanging fruit of RL on various job tasks... but how does that work when the job, the tasks and the tools are in a state of flux?
|
after finishing the recent Dwarkesh podcast with the Anthropic guys, i don't think they mentioned continual learning once (or at least, no time was spent on the subject). i think they did touch on long-term memory but only in the context of text, no visual or muscle memory.
| 1 |
2025-05-24T20:19:26.869Z
|
2025-05-24T20:19:26.817Z
|
gracekind.net
|
did:plc:p572wxnsuoogcrhlfrlizlrb
|
at://did:plc:p572wxnsuoogcrhlfrlizlrb/app.bsky.feed.post/3lpufzvpmt22a
|
bafyreieaxmv5tckqvr6xnqsg2s233ymrlvqykgqh34vnzgupsosq4r7vvq
|
It’s weird because this really doesn’t seem like news to me. Opus 4 is a big deal but this particular behavior is a drop in the ocean. I’m bad at predicting what goes viral, though And I forget that most peoples’ experience with AI is “the google AI summary was wrong again”
|
It’s weird because this really doesn’t seem like news to me. Opus 4 is a big deal but this particular behavior is a drop in the ocean. I’m bad at predicting what goes viral, though
| 1 |
2025-05-23T19:56:50.100Z
|
2025-05-23T19:56:50.813Z
|
keytryer.bsky.social
|
did:plc:u5stolpomsfbkgqfopef6jio
|
at://did:plc:u5stolpomsfbkgqfopef6jio/app.bsky.feed.post/3lqt2gpr3z22x
|
bafyreib77b2zsjfi25ag5tkheub3ed5mm7hlnezbdbvpv2pzbjfqd5hpqm
|
It may sound like a massive epistemic difference, but the anti-AI posters on this site legitimately believe that LLMs are magic tricks, they either don't exist or don't work (And will go away any time soon), and so they think they'll have 0 value compared to anything.
|
It may sound like a massive epistemic difference, but the anti-AI posters on this site legitimately believe that LLMs are magic tricks, they either don't exist or don't work (And will go away any time soon), and so they think they'll have 0 value compared to anything.
| 0 |
2025-06-05T00:21:47.052Z
|
2025-06-05T00:21:47.720Z
|
natolambert.bsky.social
|
did:plc:brkj2yocng7vtggmyujy4khq
|
at://did:plc:brkj2yocng7vtggmyujy4khq/app.bsky.feed.post/3lkm6hdfmas2l
|
bafyreian6jn7vcysfrjutkqxwyxpw2f3l557ypdytvfilpk3pu5vkcupxa
|
3+, see discussion on X :/ x.com/danielhanche... 4. Reward shaping for truncation
TLDR is a reward shaping for samples that are too long and distort learning signal. This is very logical.
|
3+, see discussion on X :/ x.com/danielhanche...
| 1 |
2025-03-17T22:13:48.664Z
|
2025-03-17T22:13:57.890Z
|
tedunderwood.me
|
did:plc:565ebob5f6hw33hjdkxty6qj
|
at://did:plc:565ebob5f6hw33hjdkxty6qj/app.bsky.feed.post/3lquqhn363c26
|
bafyreibn3y5fbaog7z5xj3g7vq5gcaydenlqs2b2cjl54dpeor5ro2273u
|
Two pieces of news here. One is, new chancellor! The other is, the Univ of Illinois System is on Bluesky. I think good news in both cases.
|
Two pieces of news here. One is, new chancellor! The other is, the Univ of Illinois System is on Bluesky.
| 1 |
2025-06-05T16:28:39.850Z
|
2025-06-05T16:28:40.113Z
|
tachikoma.elsewhereunbound.com
|
did:plc:mpdezz4nkre7vyift2rttggl
|
at://did:plc:mpdezz4nkre7vyift2rttggl/app.bsky.feed.post/3lpzo7hnji22p
|
bafyreic5vjglvrlihrzkk5g6toswnqsaxswrexsie674mx7vokeir2lefq
|
good to know, i will look that up. so far i've just kept the same terminal session open, but i just started today.
|
good to know, i will look that up. so far i've just kept the same terminal session open, but i just started today.
| 0 |
2025-05-25T22:06:25.304Z
|
2025-05-25T22:06:25.715Z
|
tachikoma.elsewhereunbound.com
|
did:plc:mpdezz4nkre7vyift2rttggl
|
at://did:plc:mpdezz4nkre7vyift2rttggl/app.bsky.feed.post/3lr2a2ckp3c24
|
bafyreigjv4ohc3upl6sd46tqou6kzysgxh3hooyvg62vzohrb7z7zozjca
|
what if in waiting we let less enlightened groups choke off the exists such that you must pass through their enclosures where they can trap you?
|
what if in waiting we let less enlightened groups choke off the exists such that you must pass through their enclosures where they can trap you?
| 0 |
2025-06-07T20:50:50.971Z
|
2025-06-07T20:50:51.342Z
|
deepfates.com.deepfates.com.deepfates.com.deepfates.com.deepfates.com
|
did:plc:7ti53c7iwv3poxnypw2tu3ck
|
at://did:plc:7ti53c7iwv3poxnypw2tu3ck/app.bsky.feed.post/3lqqolhcp5c2d
|
bafyreifk6gkaux3kmzqgrwtz7xofidam7ux4jxs7ddtjzktztrl4epvjla
|
feel like everyone is sleeping on this paper? The guy who coined the bitter lesson says it's the era of experience and self-play data will vastly exceed human generated data soon? Y'all read this or..?
|
feel like everyone is sleeping on this paper? The guy who coined the bitter lesson says it's the era of experience and self-play data will vastly exceed human generated data soon? Y'all read this or..?
| 0 |
2025-06-04T01:44:21.562Z
|
2025-06-04T01:44:24.614Z
|
timkellogg.me
|
did:plc:ckaz32jwl6t2cno6fmuw2nhn
|
at://did:plc:ckaz32jwl6t2cno6fmuw2nhn/app.bsky.feed.post/3lqkulcyscs2r
|
bafyreicfoy5i5rznzc7iumvlolzxdt3zx52ns3i6s5wk3tnuufc5qvjqze
|
this pic is going around and backend devs are all like, “this is amazing!”
c’mon, take a second and read the code
1. PUT on a customer ID schedules a delivery? really?
2. DELETE customer ID prints a receipt???!
it just keeps getting worse the longer you look
|
this pic is going around and backend devs are all like, “this is amazing!”
c’mon, take a second and read the code
1. PUT on a customer ID schedules a delivery? really?
2. DELETE customer ID prints a receipt???!
it just keeps getting worse the longer you look
| 0 |
2025-06-01T18:15:41.080Z
|
2025-06-01T18:15:46.111Z
|
keytryer.bsky.social
|
did:plc:u5stolpomsfbkgqfopef6jio
|
at://did:plc:u5stolpomsfbkgqfopef6jio/app.bsky.feed.post/3lqsftcqink2o
|
bafyreiestmkbndq4df63hd2ik2dek5fha5jbd5mqon5qfxmcektqkmogua
|
I think this says a lot about how confident they are in winning those copyright infringement lawsuits
www.musicbusinessworldwide.com/universal-wa...
|
I think this says a lot about how confident they are in winning those copyright infringement lawsuits
www.musicbusinessworldwide.com/universal-wa...
| 0 |
2025-06-04T18:13:01.030Z
|
2025-06-04T18:13:05.623Z
|
tachikoma.elsewhereunbound.com
|
did:plc:mpdezz4nkre7vyift2rttggl
|
at://did:plc:mpdezz4nkre7vyift2rttggl/app.bsky.feed.post/3lqatuaewds2b
|
bafyreib4fyrp3fepte3hlpcbrpsyuqktbg2mlxlrhpnm7aeo44fjeufqi4
|
this is a bit tough, even humans do brute force search from time to time and call it thinking, like finding the path through a complex maze. the difference is that we can go beyond search to pattern matching (and beyond that too) which computers only matched recently with ANNs.
|
this is a bit tough, even humans do brute force search from time to time and call it thinking, like finding the path through a complex maze. the difference is that we can go beyond search to pattern matching (and beyond that too) which computers only matched recently with ANNs.
| 0 |
2025-05-28T18:36:09.204Z
|
2025-05-28T18:36:09.113Z
|
nmk.wtf
|
did:plc:xm55llyxcrb3h7qwqr3wtaog
|
at://did:plc:xm55llyxcrb3h7qwqr3wtaog/app.bsky.feed.post/3lob5srnzqc2h
|
bafyreif7mcfmypiljotnzwxayd54h7o4x2n755ynlopnd7hy4lfa7gsgqy
|
I wrote a post on a Harberger Tax on patents (with a focus on pharmaceutical companies)
|
I wrote a post on a Harberger Tax on patents (with a focus on pharmaceutical companies)
| 0 |
2025-05-03T10:43:54.380Z
|
2025-05-03T10:43:56.750Z
|
gracekind.net
|
did:plc:p572wxnsuoogcrhlfrlizlrb
|
at://did:plc:p572wxnsuoogcrhlfrlizlrb/app.bsky.feed.post/3lq6jwd2etc2c
|
bafyreiev7iemn7d5ldzghvypkx2oqasewmnw3byo2j2madyvdaqxev72nq
|
We put 10 skeptics in a room. They’ll never believe what happens next
|
We put 10 skeptics in a room. They’ll never believe what happens next
| 0 |
2025-05-27T20:33:02.218Z
|
2025-05-27T20:33:02.811Z
|
norvid-studies.bsky.social
|
did:plc:mssgex5rqek4wc66wgvzztbc
|
at://did:plc:mssgex5rqek4wc66wgvzztbc/app.bsky.feed.post/3lqn5dawx6s2c
|
bafyreicfx427q5k3p24buchjd2ookuvkntax7lp4gq6sfpmxzmmzoaqlyi
|
"kelsey piper and the rats!" sort of indie band naming scheme
|
"kelsey piper and the rats!" sort of indie band naming scheme
| 0 |
2025-06-02T15:57:33.648Z
|
2025-06-02T15:57:34.013Z
|
segyges.bsky.social
|
did:plc:vszw3ess46odfhnzdsy4huae
|
at://did:plc:vszw3ess46odfhnzdsy4huae/app.bsky.feed.post/3lqvhccf22222
|
bafyreic4ou27q3u7oitc6xqa2bcuinejrygjvbxpnn5u7k4zm4brfn6mwq
|
we told a lawyer that it was his job to defend this theory for why elon should keep control of spacex in court and he killed himself
|
we told a lawyer that it was his job to defend this theory for why elon should keep control of spacex in court and he killed himself
| 0 |
2025-06-05T23:17:16.929Z
|
2025-06-05T23:17:17.712Z
|
gracekind.net
|
did:plc:p572wxnsuoogcrhlfrlizlrb
|
at://did:plc:p572wxnsuoogcrhlfrlizlrb/app.bsky.feed.post/3lq7hnzbcyk25
|
bafyreih6aifnl7bbgs5dnooo36eod3livsoqszjnv2tzhn2twcpeb34cu4
|
I read this as “people familiar with plumbing” which would also make sense!
|
I read this as “people familiar with plumbing” which would also make sense!
| 0 |
2025-05-28T05:25:15.779Z
|
2025-05-28T05:25:16.811Z
|
defenderofbasic.bsky.social
|
did:plc:sydtqdji6dxfqfn35gd7esdw
|
at://did:plc:sydtqdji6dxfqfn35gd7esdw/app.bsky.feed.post/3ldj5ts2dh22h
|
bafyreifmzj4nezgvmhhi3ruyyw27w3izzp6lsrgctsqwxv5odwasxjvhxa
|
my search page shows me 3d ago, and 18d ago. It's so much missing stuff none of my interactions with @lemonaut1.bsky.social are in there for example
|
my search page shows me 3d ago, and 18d ago. It's so much missing stuff
| 1 |
2024-12-17T15:28:38.340Z
|
2024-12-17T15:28:40.717Z
|
natolambert.bsky.social
|
did:plc:brkj2yocng7vtggmyujy4khq
|
at://did:plc:brkj2yocng7vtggmyujy4khq/app.bsky.feed.post/3lgetmftgzz2g
|
bafyreigghxprhnfcaohsz2gdr3ybcqibrdexrwk6wjaqqzp7yzjuunttpa
|
Change my mind: The safety work on o1 from OpenAI is already a ton more evidence than people are giving credit for that reasoning-heavy training will generalize benefits to other domains.
Safety is super orthogonal from math/code! I expect many more examples by end of 2025.
|
Change my mind: The safety work on o1 from OpenAI is already a ton more evidence than people are giving credit for that reasoning-heavy training will generalize benefits to other domains.
Safety is super orthogonal from math/code! I expect many more examples by end of 2025.
| 0 |
2025-01-23T02:30:29.045Z
|
2025-01-23T02:30:29.130Z
|
emollick.bsky.social
|
did:plc:flxq4uyjfotciovpw3x3fxnu
|
at://did:plc:flxq4uyjfotciovpw3x3fxnu/app.bsky.feed.post/3lo63kzn5ys2b
|
bafyreichs3bk5c4pyf73233wkdek72xauzqd22u5ij4slhvczyds3pnwrq
|
“Make a linked in page (as an image) for business cyber duck”
|
“Make a linked in page (as an image) for business cyber duck”
| 0 |
2025-05-02T05:25:47.599Z
|
2025-05-02T05:25:48.849Z
|
gracekind.net
|
did:plc:p572wxnsuoogcrhlfrlizlrb
|
at://did:plc:p572wxnsuoogcrhlfrlizlrb/app.bsky.feed.post/3lqxc7zxzv224
|
bafyreibd4gnhbwcql5hcrpbcyhbguirjwngxrtjboudpegrlztiijvhaz4
|
I agree that compression is a measure of intelligence- I guess I don’t have an intuition for how the Turing test maps to data compression
|
I agree that compression is a measure of intelligence- I guess I don’t have an intuition for how the Turing test maps to data compression
| 0 |
2025-06-06T16:51:51.772Z
|
2025-06-06T16:51:52.616Z
|
natolambert.bsky.social
|
did:plc:brkj2yocng7vtggmyujy4khq
|
at://did:plc:brkj2yocng7vtggmyujy4khq/app.bsky.feed.post/3lg6pkyaky22x
|
bafyreibsizx7burbijlkgoc6yzhnil54vnjtbyt3m6oxybfsqvxeqb24b4
|
Oh yeah you’re right ima delete my tweet. I should just write long form
|
Oh yeah you’re right ima delete my tweet. I should just write long form
| 0 |
2025-01-20T16:02:07.960Z
|
2025-01-20T16:02:08.617Z
|
defenderofbasic.bsky.social
|
did:plc:sydtqdji6dxfqfn35gd7esdw
|
at://did:plc:sydtqdji6dxfqfn35gd7esdw/app.bsky.feed.post/3ldtacumlhs27
|
bafyreic5k2eqfrza5o4g2noxczjbkhociqf7xqvgwxt3sx6fut7gwu2nzu
|
Ah I think I don't understand your point then! I thought you were unironically saying "it's silly to be upset at people saying words on social media, because it has no consequence"
|
Ah I think I don't understand your point then! I thought you were unironically saying "it's silly to be upset at people saying words on social media, because it has no consequence"
| 0 |
2024-12-21T15:39:29.220Z
|
2024-12-21T15:39:29.918Z
|
deepfates.com.deepfates.com.deepfates.com.deepfates.com.deepfates.com
|
did:plc:7ti53c7iwv3poxnypw2tu3ck
|
at://did:plc:7ti53c7iwv3poxnypw2tu3ck/app.bsky.feed.post/3lbxomxrrxs2e
|
bafyreihohi3kxxd6zkelj3jkd2kyb2inw3namq2ovzaf5mo26nvvklamdi
|
The other option was have no users. I understand why they've made the calls they have. The core team doesn't like these people any more than you or i do, but they have to put up with them because the point is to actually change the internet
|
The other option was have no users. I understand why they've made the calls they have. The core team doesn't like these people any more than you or i do, but they have to put up with them because the point is to actually change the internet
| 0 |
2024-11-27T23:15:56.163Z
|
2024-11-27T23:15:55.523Z
|
ens0.me
|
did:plc:623st67kkthivj4c6icvkqnq
|
at://did:plc:xfjlu7iu6xl7uyi4wybbgx3z/app.bsky.feed.post/3lqi7kx2e322f
|
bafyreig4v6ghfzmxfczqg6ylfajmic57iig6chlpsjcyt4bvvbphushc4a
|
Completely fair to disparage their drug and alcohol use, when normal people get excluded from federal service for the exact same (or even more benign) behaviors tbh
|
Completely fair to disparage their drug and alcohol use, when normal people get excluded from federal service for the exact same (or even more benign) behaviors tbh
| 0 |
2025-05-31T16:54:20.500Z
|
2025-05-31T16:54:22.013Z
|
nsaphra.bsky.social
|
did:plc:l3vbx3e2lqnlja7q64wfrrzq
|
at://did:plc:l3vbx3e2lqnlja7q64wfrrzq/app.bsky.feed.post/3lqvl6df4tc23
|
bafyreidzgqyinzuhnrfopaqp7pnh4wkmlnmvmoppmns45zbrkjuhevgdvi
|
If the federal government collapses do you think we can get cancer research grants again
|
If the federal government collapses do you think we can get cancer research grants again
| 0 |
2025-06-06T00:26:38.728Z
|
2025-06-06T00:26:39.215Z
|
timkellogg.me
|
did:plc:ckaz32jwl6t2cno6fmuw2nhn
|
at://did:plc:ckaz32jwl6t2cno6fmuw2nhn/app.bsky.feed.post/3lpyzx34ckc2g
|
bafyreia4scyiqfucnol3ivhel3ld3osaurwimelzn6lwmenpjv7ccdwylm
|
depends on how much info is in the picture, but not very often
|
depends on how much info is in the picture, but not very often
| 0 |
2025-05-25T16:03:48.886Z
|
2025-05-25T16:03:49.730Z
|
segyges.bsky.social
|
did:plc:vszw3ess46odfhnzdsy4huae
|
at://did:plc:vszw3ess46odfhnzdsy4huae/app.bsky.feed.post/3lqxljfhr6224
|
bafyreie7jpzdw74rjytl3qpcl3vlmm4p7lcnvcs5mw6zkkazivxtapuk2e
|
nor do we have one that is likely to make errors at a human-like rate: his test has an arithmetic error, and so "making human-like errors at a human-like rate" is plausibly a requirement, too fml this becomes a longform article i guess
|
nor do we have one that is likely to make errors at a human-like rate: his test has an arithmetic error, and so "making human-like errors at a human-like rate" is plausibly a requirement, too
| 1 |
2025-06-06T19:38:09.489Z
|
2025-06-06T19:38:10.119Z
|
tachikoma.elsewhereunbound.com
|
did:plc:mpdezz4nkre7vyift2rttggl
|
at://did:plc:mpdezz4nkre7vyift2rttggl/app.bsky.feed.post/3lqh4dsjdcs2z
|
bafyreicdg5wmetkkqepaqzot34jn5xel7ostdxttfec6vzywnzduc7okcm
|
i meant Vinge wasn't an AI researcher, but then most of the people on that list aren't either. also i'm not one either.
on the robotics side, there's the concept of the uncanny valley. maybe a mention of the Eliza effect?
|
i meant Vinge wasn't an AI researcher, but then most of the people on that list aren't either. also i'm not one either.
on the robotics side, there's the concept of the uncanny valley. maybe a mention of the Eliza effect?
| 0 |
2025-05-31T06:23:59.904Z
|
2025-05-31T06:24:00.211Z
|
tedunderwood.me
|
did:plc:565ebob5f6hw33hjdkxty6qj
|
at://did:plc:j335z25k4etozod73flbnlpt/app.bsky.feed.post/3lqmlfdrepk2o
|
bafyreiethnz2o5a4i5iszjxttkowhbxzn4xpdemp2jiofhja5mllgxslji
|
I am excited to announce our latest work 🎉 "Cultural Evaluations of Vision-Language Models Have a Lot to Learn from Cultural Theory". We review recent works on culture in VLMs and argue for deeper grounding in cultural theory to enable more inclusive evaluations.
Paper 🔗: arxiv.org/pdf/2505.22793
|
I am excited to announce our latest work 🎉 "Cultural Evaluations of Vision-Language Models Have a Lot to Learn from Cultural Theory". We review recent works on culture in VLMs and argue for deeper grounding in cultural theory to enable more inclusive evaluations.
Paper 🔗: arxiv.org/pdf/2505.22793
| 0 |
2025-06-02T10:36:36.359Z
|
2025-06-02T10:36:45.317Z
|
tachikoma.elsewhereunbound.com
|
did:plc:mpdezz4nkre7vyift2rttggl
|
at://did:plc:mpdezz4nkre7vyift2rttggl/app.bsky.feed.post/3lpvink7gwk2q
|
bafyreid455v5oudavlxpqzjal4v5bee7mwwbbemyafygwcjpesrv5rarzy
|
so much discussion around AI rests on reliability, and very few are willing to update all the way and ask what should the norms be if the models are indeed as reliable if not more than a human. i told my coworkers in a tech field and living just by SF about my trip in a Waymo last week and they all seemed gobsmacked by it. you live there! they aren't even that new anymore! how have you not updated yet?
|
so much discussion around AI rests on reliability, and very few are willing to update all the way and ask what should the norms be if the models are indeed as reliable if not more than a human.
| 1 |
2025-05-24T06:16:16.346Z
|
2025-05-24T06:16:13.414Z
|
ens0.me
|
did:plc:623st67kkthivj4c6icvkqnq
|
at://did:plc:623st67kkthivj4c6icvkqnq/app.bsky.feed.post/3lqpvwfexfs27
|
bafyreiaq6m6si3bjf7m2e75zc2dkttzrnwjhe6uhoyg36nepc4lsbtmh6q
|
When I was in middle/high school, it was a big deal you could just MapQuest directions and print them out 😭
|
When I was in middle/high school, it was a big deal you could just MapQuest directions and print them out 😭
| 0 |
2025-06-03T18:23:05.116Z
|
2025-06-03T18:23:05.344Z
|
tachikoma.elsewhereunbound.com
|
did:plc:mpdezz4nkre7vyift2rttggl
|
at://did:plc:mpdezz4nkre7vyift2rttggl/app.bsky.feed.post/3lqfw3k75j22n
|
bafyreifoisjcrngmu6g3be4exvqajflubg72k7hnsrzrivc2muuqwiyvbq
|
it's a bit annoying that stuff like this doesn't show up on my TL, whether following or discover. i don't really like having to flick between half-a-dozen or more feeds to keep up with various subjects... spooky. also a simple reward to reduce tool use hallucination is... well crude. but this is all research i suppose.
|
it's a bit annoying that stuff like this doesn't show up on my TL, whether following or discover. i don't really like having to flick between half-a-dozen or more feeds to keep up with various subjects...
| 1 |
2025-05-30T18:59:20.554Z
|
2025-05-30T18:59:19.813Z
|
emollick.bsky.social
|
did:plc:flxq4uyjfotciovpw3x3fxnu
|
at://did:plc:flxq4uyjfotciovpw3x3fxnu/app.bsky.feed.post/3lmmxzqiz7s2j
|
bafyreihwberjb6reb3oosa3n6du7i6ht725uon4snweh7m5kbbpiwfvya4
|
Anthropic continues to put out solid studies of how Claude is used, this time in universities
It shows both AI perils & promise. Students are getting direct answers from Claude, making up a lot use, but there are many cases of AI supporting their learning; too. www.anthropic.com/news/anthrop... I think the interpretation of these results and what they mean are not completely clear, but what is clear that AI use is already deeply a part of student life, which means we need to make choices about how it should be used.
|
Anthropic continues to put out solid studies of how Claude is used, this time in universities
It shows both AI perils & promise. Students are getting direct answers from Claude, making up a lot use, but there are many cases of AI supporting their learning; too. www.anthropic.com/news/anthrop...
| 1 |
2025-04-12T16:41:59.195Z
|
2025-04-12T16:42:05.395Z
|
ens0.me
|
did:plc:623st67kkthivj4c6icvkqnq
|
at://did:plc:623st67kkthivj4c6icvkqnq/app.bsky.feed.post/3lqxgxnqork2l
|
bafyreibx27ph4bprhjglkicbup5qiere5va6rv2twjynkylepx3auh7kiu
|
Yeah, the whole exaggeration for effect versus outright lying thing.
Like I agree it wouldn't be in character for Adam to say that pasta water should be as salty as the Mediterranean. That would be too salty.
But that's something someone would just say to be cute.
|
Yeah, the whole exaggeration for effect versus outright lying thing.
Like I agree it wouldn't be in character for Adam to say that pasta water should be as salty as the Mediterranean. That would be too salty.
But that's something someone would just say to be cute.
| 0 |
2025-06-06T18:16:39.223Z
|
2025-06-06T18:16:39.715Z
|
nsaphra.bsky.social
|
did:plc:l3vbx3e2lqnlja7q64wfrrzq
|
at://did:plc:l3vbx3e2lqnlja7q64wfrrzq/app.bsky.feed.post/3lmpc2ogwks2t
|
bafyreidsffjgq7tizuyc3lkrntqxpnst5odp7iugserzltxm3bm5zs2ata
|
Yeah, it is literally impossible to capture every recent tangentially related flag plant and still have a readable paper.
|
Yeah, it is literally impossible to capture every recent tangentially related flag plant and still have a readable paper.
| 0 |
2025-04-13T14:46:47.486Z
|
2025-04-13T14:46:47.947Z
|
ens0.me
|
did:plc:623st67kkthivj4c6icvkqnq
|
at://did:plc:blpxxtqvnvfvyp7fo5kibzrw/app.bsky.feed.post/3lquoehzh7s2l
|
bafyreifh6xlos76cz5lk6pescalhct2pusksj4eqo6zt4fafmndo3jkeya
|
Also, in an era where climate change will almost certainly affect crop yields and where crops can be grown, agricultural research becomes even more important to ensure we can continue to feed people
|
Also, in an era where climate change will almost certainly affect crop yields and where crops can be grown, agricultural research becomes even more important to ensure we can continue to feed people
| 0 |
2025-06-05T15:51:06.401Z
|
2025-06-05T15:51:06.817Z
|
tachikoma.elsewhereunbound.com
|
did:plc:mpdezz4nkre7vyift2rttggl
|
at://did:plc:mpdezz4nkre7vyift2rttggl/app.bsky.feed.post/3lqxbgb6das2g
|
bafyreihlcfpcr4uci7vgnma7gt7jazfc5qoctribhffkko6yyubhd2vpq4
|
an artifact of how the test is constructed, like other academic tests that LLMs ace. it's too short. intelligence has to be measured over time. the delta, learning in particular, is the key.
|
an artifact of how the test is constructed, like other academic tests that LLMs ace. it's too short. intelligence has to be measured over time. the delta, learning in particular, is the key.
| 0 |
2025-06-06T16:37:26.904Z
|
2025-06-06T16:37:26.113Z
|
segyges.bsky.social
|
did:plc:vszw3ess46odfhnzdsy4huae
|
at://did:plc:vszw3ess46odfhnzdsy4huae/app.bsky.feed.post/3lqvvyfof3s2b
|
bafyreidafdu7r4u5n4fjkfagzmt7nditshvjnooqcpfxdzokrtph2dpuci
|
my worst scenarios are all bad but not so bad that anything changes, yeah. in that goldilocks zone worse is better. once you're past it faster is just more dangerous tho we already built a panopticon and god knows what else we'll make
|
my worst scenarios are all bad but not so bad that anything changes, yeah. in that goldilocks zone worse is better. once you're past it faster is just more dangerous tho
| 1 |
2025-06-06T03:40:10.964Z
|
2025-06-06T03:40:11.513Z
|
deepfates.com.deepfates.com.deepfates.com.deepfates.com.deepfates.com
|
did:plc:7ti53c7iwv3poxnypw2tu3ck
|
at://did:plc:7ti53c7iwv3poxnypw2tu3ck/app.bsky.feed.post/3lpp3tpzxys2k
|
bafyreigy6y3xksbspbnyvchecsxhooxp7ohmxuhvc7vwadbvz7gw7e2rhi
|
Nice. O did cross post for a while but i find it even more addictive than just one or the other. Like those cards that say "how to keep an idiot busy (flip over)" on both sides
|
Nice. O did cross post for a while but i find it even more addictive than just one or the other. Like those cards that say "how to keep an idiot busy (flip over)" on both sides
| 0 |
2025-05-21T17:11:06.711Z
|
2025-05-21T17:11:06.415Z
|
norvid-studies.bsky.social
|
did:plc:mssgex5rqek4wc66wgvzztbc
|
at://did:plc:mssgex5rqek4wc66wgvzztbc/app.bsky.feed.post/3lqxfoikqwc2v
|
bafyreihy7ufn7kkbzmvgvqqy666ofqn7oln7ygbxs5cvh5ituu4huavvki
|
and reforms so it just has less room but the remaining tasks are "denser" because "complementarily more valuable" (here the physical metaphor ceases to function)-- for a while. but then eventually it just gets pushed off the circle altogether, there's nothing left in the range.
|
and reforms so it just has less room but the remaining tasks are "denser" because "complementarily more valuable" (here the physical metaphor ceases to function)-- for a while. but then eventually it just gets pushed off the circle altogether, there's nothing left in the range.
| 0 |
2025-06-06T17:53:38.053Z
|
2025-06-06T17:53:38.312Z
|
ens0.me
|
did:plc:623st67kkthivj4c6icvkqnq
|
at://did:plc:623st67kkthivj4c6icvkqnq/app.bsky.feed.post/3lqxgpcfh722l
|
bafyreighmk7klaz57hr7xhw4ofbcbpnero5nzpbbi3zoppm2artzqz5tae
|
People want a lot of things, and I think the bounds of what they'll want will expand in this sci-fi future into which we're entering — though a bit too cyberpunk dystopia for my tastes on its current trajectory.
But I do think a lot of people crave human authenticity. I use it extensively to help me code now, where it has gotten VERY good, but I still write every word by hand — I'm in the NYT years ago emdashing it up so haters can stay mad that I'm committing that apparent faux pas.
Because I like what I write being authentic to me.
|
People want a lot of things, and I think the bounds of what they'll want will expand in this sci-fi future into which we're entering — though a bit too cyberpunk dystopia for my tastes on its current trajectory.
But I do think a lot of people crave human authenticity.
| 1 |
2025-06-06T18:11:58.885Z
|
2025-06-06T18:11:59.314Z
|
timkellogg.me
|
did:plc:ckaz32jwl6t2cno6fmuw2nhn
|
at://did:plc:ckaz32jwl6t2cno6fmuw2nhn/app.bsky.feed.post/3lr422gxuuc2a
|
bafyreidjt5z7m3cx7qqzggcbxy7cejqkgqwokosnurncousjsmrece6ara
|
if it’s hierarchical it could make sense as a resource, bc then the exploration aspect works. but if it’s relational then it’s only useful as a cache key
|
if it’s hierarchical it could make sense as a resource, bc then the exploration aspect works. but if it’s relational then it’s only useful as a cache key
| 0 |
2025-06-08T14:08:52.829Z
|
2025-06-08T14:08:53.319Z
|
timkellogg.me
|
did:plc:ckaz32jwl6t2cno6fmuw2nhn
|
at://did:plc:ckaz32jwl6t2cno6fmuw2nhn/app.bsky.feed.post/3lqhpmwmizk2a
|
bafyreicinwml3c343krw6myohwhrpiyn2vsipkghthsct47mowlqnxfy7q
|
prompts:
- while generating a prompt, load new tools and use the returned prompt to instruct how to use them
- use an LLM to format a prompt
- lookup info from a database and pump it into a prompt resources:
- tool accepts a resource URI (e.g. a DB record)
- tool returns 3 URIs
- prompt references URIs (e.g. a base directory to work from)
- proactive pushing resources (always do this if compatibility is important)
- reactive pulling resources (do this if network bandwidth is important)
|
prompts:
- while generating a prompt, load new tools and use the returned prompt to instruct how to use them
- use an LLM to format a prompt
- lookup info from a database and pump it into a prompt
| 1 |
2025-05-31T12:09:07.286Z
|
2025-05-31T12:09:08.413Z
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.