Rikard Saqe

1 followers · 2058 views

on the atlas · 337

highlights · 772

Any singular superintelligence would have to prioritize some values over others and in the process would be incapable of being benevolent to everyone.
wsj.com
When people are empowered, they naturally check and balance each other and the power of larger institutions.
wsj.com
delivering personal superintelligence to everyone
wsj.com
the best approach is letting people decide what matters to them
wsj.com
the values that got us to this point—liberty, open inquiry, free enterprise and equal opportunity—are also the values to build a positive future.
wsj.com
shift scientific publishing practices
NSF announces $400M investment in new national network of AI-programmable cloud laboratories in alignment with the U.S. government's Genesis Mission | NSF - U.S. National Science Foundation
achieving a virtuous cycle of automated hypothesis generation, autonomous experimentation and generation and interpretation of large volumes of high-quality experimental data to drive the next set of hypotheses, all with researchers and students alongside the automated systems at every step of the way.
NSF announces $400M investment in new national network of AI-programmable cloud laboratories in alignment with the U.S. government's Genesis Mission | NSF - U.S. National Science Foundation
The NSF PCL Test Bed program directly implements a priority of the White House AI Action Plan as well as a recommendation of the bipartisan National Security Commission on Emerging Biotechnology to accelerate AI-enabled science through automated laboratory infrastructure.
NSF announces $400M investment in new national network of AI-programmable cloud laboratories in alignment with the U.S. government's Genesis Mission | NSF - U.S. National Science Foundation
prioritize open science, reusability and reproducibility, AI-readiness of data, and the rapid publication of data and results
NSF announces $400M investment in new national network of AI-programmable cloud laboratories in alignment with the U.S. government's Genesis Mission | NSF - U.S. National Science Foundation
Decoupling food from land accomplishes little if we recouple it to industrial sugar.
Scaling alternative proteins for net benefits: One Earth
We should engineer microbial hosts to metabolize one-carbon and waste feedstocks, including CO₂, methane, formate, and food-industry sidestreams, instead of first-generation sugars, while diversifying our chassis repertoire rather than standardizing around a few strains.
Scaling alternative proteins for net benefits: One Earth
As the field converges on a handful of workhorse chassis, a few sugar sources, and centralized mega-biofoundries, we risk reconstructing industrial agriculture’s defining pathologies: input dependency, genetic uniformity, and corporate concentration, one trophic level downward.
Scaling alternative proteins for net benefits: One Earth
A more resilient pathway depends on diversification rather than substitution. Legume-based systems illustrate this potential: pulses can support low-input agriculture, enhance soil fertility through biological nitrogen fixation, and diversify farming landscapes.
Scaling alternative proteins for net benefits: One Earth
artificial intelligence (AI) could enable high-throughput analysis of protein sequences, epitope conservatism and cross-species homology to forecast allergenicity for whole food chains, covering raw material access, production and processing, and market circulation. Standardized raw-material screening aligned with international safety standards would be the first stage. AI could guide the optimization of desensitizing processes including enzymatic hydrolysis, non-thermal process, and fermentation to disrupt allergen structures while preserving nutrition.
Scaling alternative proteins for net benefits: One Earth
agenda setting would include (a) the identification of an ambitious but achievable goal; (b) the specific actions that would need to be taken to achieve the goal; and (c) the rationale for philanthropic involvement.
Kalil’s Corner: The Role of Philanthropy in Identifying and Pursuing AI Moonshots
agenda setting would include (a) the identification of an ambitious but achievable goal; (b) the specific actions that would need to be taken to achieve the goal; and (c) the rationale for philanthropic involvement.
Kalil’s Corner: The Role of Philanthropy in Identifying and Pursuing AI Moonshots
agenda setting would include (a) the identification of an ambitious but achievable goal; (b) the specific actions that would need to be taken to achieve the goal; and (c) the rationale for philanthropic involvement.
Kalil’s Corner: The Role of Philanthropy in Identifying and Pursuing AI Moonshots
TheNew York of Canada” is an oxymoron.
Stop Pretending Toronto is More Than It Is
it’s the same few examples over and over. It’s Drake, Insulin, Shopify, Hinton.
Stop Pretending Toronto is More Than It Is
The firm operates the largest food-grade cell culture bioreactor at 22,000 litres, and claims to have completed the largest cultivated meat harvest in history (1,500 kg). Currently, it is producing over 5,000 kg of cultivated meat per week.
Australia's Vow Appoints New CEO Amid Expansion Beyond Lab-Grown Meat
Vow’s cell cultivation capacity was extended to 35,000 litres within its second factory, which it says was 20 to 50 times cheaper to build than competitors.
Australia's Vow Appoints New CEO Amid Expansion Beyond Lab-Grown Meat
If you have any young friends who aspire to become writers, the second-greatest favor you can do them is to present them with copies of The Elements of Style. The first-greatest, of course, is to shoot them now, while they're happy.
The_Elements_of_Style
conducted dozens of runs at the 2,000 liter scale
UPSIDE Foods submits $50m bid for Believer Meats’ US assets
The change in our image of leaders and experts leaves us with”—and this is exactly your point—“a distrust of power, but also with a seemingly powerless dependence on those in whom we have little trust.”
The Orality Theory of Everything - The Atlantic
Future work should focus on standardising PPH production for consistent bioactivity and scalability, optimising formulations through defined PPH compositions or combinations with growth factors to achieve cost-effective serum-free media.
Sustainable food-grade serum-reducing biomaterial: Plant protein hydrolysates for scalable cultivated meat manufacturing - ScienceDirect
Working on AI text, as an editor, is like trying to operate on a body whose skin, muscles, veins, bones, and organs are all compromised. There’s nothing to leave intact, nowhere to begin.
The Biggest Tell That Something Was Written by AI - The Atlantic
AI writing is almost impossible to edit, because even when it sounds plausible, a closer look will show that every element is equally off: The tone is bland; individual word choices are baffling; the structure lacks sense; key pieces of the argument are missing; facts are false.
The Biggest Tell That Something Was Written by AI - The Atlantic
we end up with canned perfection—writing that can’t really be argued with, because it has no underlying deliberative reasoning process, no train of thought
The Biggest Tell That Something Was Written by AI - The Atlantic
a project whose hubris is underrated
The Biggest Tell That Something Was Written by AI - The Atlantic
Many authors describe how, when they’ve finally hit on the right idea, writing feels like going down a water slide; putting one sentence after another becomes easy.
The Biggest Tell That Something Was Written by AI - The Atlantic
When we use AI to flesh out ideas, we lose the most important part of the writing process: thinking.
The Biggest Tell That Something Was Written by AI - The Atlantic
The phrase is protean and euphemistic
The Biggest Tell That Something Was Written by AI - The Atlantic
irst, an “intra-method” technique selects the best semantic match between query table and data lake table for each source (KB and synthesized KB) separately. Then an “inter-method” comparison selects the semantic match that maximizes the overall unionability score
SANTOS: Relationship-based Semantic Table Union Search
But when Pandolfi and his colleagues described the gene's role in the development of human cancer in Nature last January, the discovery attracted headlines such as ‘Pokemon's cancer role revealed’ (T. Maeda et al. Nature 433, 278–285; 2005). Message boards and blogs picked up the story, unable to resist using the phrase ‘Pokemon causes cancer’.
Pokémon blocks gene name | Nature
A spokeswoman for Pokémon USA told Nature that its image was at risk. “We don't want our image undermined by associating Pokémon with cancer,” she said.
Pokémon blocks gene name | Nature
Following the same logic as annotating synthesized CS, we consider two column pairs to have the same RS if their value pairs overlap. For instance, column pairs (A, B) and (D, E) in Table 2 of Fig. 2 have overlapping value pairs (depicted in bold text), indicating that they likely share a meaningful relationship.
SANTOS: Relationship-based Semantic Table Union Search
we do not have type hierarchy information for the synthesized column semantics, we assume all synthesized CS are of the same granularity level and set 𝑔𝑠 (𝑎) = 1 for each type.
SANTOS: Relationship-based Semantic Table Union Search
ssuming we do not have homographs
SANTOS: Relationship-based Semantic Table Union Search
Rather than finding actual semantic types, we leverage type co-occurrence across columns to decide their semantic compatibility.
SANTOS: Relationship-based Semantic Table Union Search
e can replace the role of an existing KB in finding CS and RS with a KB that captures co-occurrence information across data lake tables.
SANTOS: Relationship-based Semantic Table Union Search
We assume that the computations of granularity score and frequency score are independent of one another as the former is based on KB statistics and the latter is based on the semantics of values in a column.
SANTOS: Relationship-based Semantic Table Union Search
We assume that the computations of granularity score and frequency score are independent of one another as the former is based on KB statistics and the latter is based on the semantics of values in a column.
SANTOS: Relationship-based Semantic Table Union Search
we will look for a tree within each semantic graph of data lake tables that matches a subtree of the query tree rooted at the intent column.
SANTOS: Relationship-based Semantic Table Union Search
specifies its intent column (denoted by 𝐼 ), which is a column of most interest to the user that forms relationships of interest with other columns. SANTOS creates a semantic graph for the query table that is restricted to being a tree rooted at the intent column, called the Query Semantic Tree. We find relationships from the intent column to other columns in the query table, then transitively search for relationships from these columns.
SANTOS: Relationship-based Semantic Table Union Search
4, an unsupervised algorithm that uses co- occurrence of values in different tables to discover the set of values that belong to the same domain. Our approach also makes use of co-occurrence of values when creating our synthesized KB, and we not only use co-occurrence of values in a column, but also consider the co-occurrence of values in binary relationships to discover relationship domains
SANTOS: Relationship-based Semantic Table Union Search
a synthesized KB to find column and relation semantics from only attribute values.
SANTOS: Relationship-based Semantic Table Union Search
Our solution determines if the binary relationship between two columns has similar semantics with other binary relationships in the data lake.
SANTOS: Relationship-based Semantic Table Union Search
We present a scoring function to match the semantic graph of a data lake table to that of a query table.
SANTOS: Relationship-based Semantic Table Union Search
Our solution creates semantic graphs for tables whose nodes are columns and edges are binary relationships found in the KB
SANTOS: Relationship-based Semantic Table Union Search
Deciding unionability based only on attribute unionability, without considering the semantics of the relationships between attributes, can lead to false positives.
SANTOS: Relationship-based Semantic Table Union Search