[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"summary-f9eef89ea9ca135d-operational-controls-beat-static-ai-governance-summary":3,"summaries-facets-categories":86,"summary-related-f9eef89ea9ca135d-operational-controls-beat-static-ai-governance-summary":3655},{"id":4,"title":5,"ai":6,"body":13,"categories":46,"created_at":48,"date_modified":48,"description":40,"extension":49,"faq":48,"featured":50,"kicker_label":48,"meta":51,"navigation":70,"path":71,"published_at":48,"question":48,"scraped_at":72,"seo":73,"sitemap":74,"source_id":75,"source_name":76,"source_type":77,"source_url":78,"stem":79,"tags":80,"thumbnail_url":48,"tldr":83,"tweet":48,"unknown_tags":84,"__hash__":85},"summaries\u002Fsummaries\u002Ff9eef89ea9ca135d-operational-controls-beat-static-ai-governance-summary.md","Operational Controls Beat Static AI Governance",{"provider":7,"model":8,"input_tokens":9,"output_tokens":10,"processing_time_ms":11,"cost_usd":12},"openrouter","x-ai\u002Fgrok-4.1-fast",14854,1633,12190,0.00376465,{"type":14,"value":15,"toc":39},"minimark",[16,21,25,29,32,36],[17,18,20],"h2",{"id":19},"distinguish-governance-policies-from-production-enforcement","Distinguish Governance Policies from Production Enforcement",[22,23,24],"p",{},"Governance sets policies, accountability, and documentation like model cards, but operational risk management enforces them in real time against deployed systems. Without this, validated models degrade silently: a fraud detection system drifted after eight months, flagging 40% more legitimate transactions because monitoring alerted only on catastrophic failures, breaching EU AI Act post-market rules for high-risk systems. Build instrumented systems to detect drift, edge cases, disparate impacts, and escalate to governance workflows—static docs alone leave compliance gaps.",[17,26,28],{"id":27},"four-core-production-ai-risks-and-mitigation-needs","Four Core Production AI Risks and Mitigation Needs",[22,30,31],{},"Address bias\u002Fdiscrimination (e.g., Workday's AI screening rejected applicants over 40, leading to a May 2025 class action), data leakage (generative models reproducing PII or inferring attributes), output risks (hallucinations like Air Canada's chatbot causing 2024 liability for false bereavement fare info), and security (prompt injection, adversarial inputs, third-party supply chain). Deploy controls for visibility: automatic event logging, anomaly detection, and incident response to make these risks observable and actionable in production.",[17,33,35],{"id":34},"nist-and-eu-ai-act-continuous-processes-over-one-time-checks","NIST and EU AI Act: Continuous Processes Over One-Time Checks",[22,37,38],{},"NIST AI RMF (Jan 2023, extended by Generative AI Profile NIST-AI-600-1 July 2024) requires ongoing Govern (accountability), Map (system inventory amid deployments\u002Ffine-tunes), Measure (quantitative risk analysis beyond pre-launch), and Manage (controls\u002Fincidents). EU AI Act (full high-risk effect Aug 2, 2026) mandates for Annex III systems (employment, credit, etc.): Article 26—monitor operations, report risks promptly, log events six months; Article 14—train overseers to detect anomalies with override authority; Article 12—build technical logging capability. Engineer these as infrastructure: dashboards for drift, human-in-loop overrides, and automated logs—not optional policies.",{"title":40,"searchDepth":41,"depth":41,"links":42},"",2,[43,44,45],{"id":19,"depth":41,"text":20},{"id":27,"depth":41,"text":28},{"id":34,"depth":41,"text":35},[47],"AI & LLMs",null,"md",false,{"content_references":52,"triage":65},[53,58,61],{"type":54,"title":55,"author":56,"publisher":56,"context":57},"report","AI Risk Management Framework","NIST","cited",{"type":59,"title":60,"context":57},"other","EU AI Act",{"type":59,"title":62,"url":63,"context":64},"The full operational implications of the EU AI Act for engineering and compliance teams","https:\u002F\u002Fsecureprivacy.ai\u002Fblog\u002Feu-ai-act-for-ctos","recommended",{"relevance":66,"novelty":67,"quality":67,"actionability":67,"composite":68,"reasoning":69},5,4,4.35,"Category: AI & LLMs. The article provides a deep dive into operational risk management for AI systems, addressing specific audience pain points such as the need for continuous monitoring and compliance with regulations like the EU AI Act. It offers actionable insights on building instrumented systems to detect risks, which is directly applicable to product builders.",true,"\u002Fsummaries\u002Ff9eef89ea9ca135d-operational-controls-beat-static-ai-governance-summary","2026-04-15 15:27:34",{"title":5,"description":40},{"loc":71},"f9eef89ea9ca135d","__oneoff__","article","https:\u002F\u002Fsecureprivacy.ai\u002Fblog\u002Foperational-ai-risk-management","summaries\u002Ff9eef89ea9ca135d-operational-controls-beat-static-ai-governance-summary",[81,82],"ai-llms","devops-cloud","AI risk management fails without continuous operational monitoring for drift, bias, and outputs—NIST and EU AI Act demand real-time logging, oversight, and escalation beyond initial docs.",[81,82],"zw9v9_DWO5dhD5iqM9uUs_FIa9NL5gCJk-Ok41Ibibk",[87,90,93,95,98,101,103,105,107,109,111,113,116,118,120,122,124,126,128,130,132,134,137,140,142,144,147,149,151,154,156,158,160,162,164,166,168,170,172,174,176,178,180,182,184,186,188,190,192,194,196,198,200,202,204,206,208,210,212,214,216,218,220,222,224,226,228,230,232,234,236,238,240,242,244,246,248,250,252,254,256,258,260,262,264,266,268,270,272,274,276,278,280,282,284,286,288,290,292,294,296,298,300,302,304,306,308,310,312,314,316,318,320,322,324,326,328,330,332,334,336,338,340,342,344,346,348,350,352,354,356,358,360,362,364,366,368,370,372,374,376,378,380,382,384,386,388,390,392,394,396,398,400,402,404,406,409,411,413,415,417,419,421,423,425,427,429,431,433,435,437,439,441,443,445,447,449,451,453,455,457,459,461,463,465,467,469,471,473,475,477,479,481,483,485,487,489,491,493,495,497,499,501,503,505,507,509,511,513,515,517,519,521,523,525,527,529,531,533,535,537,539,541,543,545,547,549,551,553,555,557,559,561,563,565,567,569,571,573,575,577,579,581,583,585,587,589,591,593,595,597,599,601,603,605,607,609,611,613,615,617,619,621,623,625,627,629,631,633,635,637,639,641,643,645,647,649,651,653,655,657,659,661,663,665,667,669,671,673,675,677,679,681,683,685,687,689,691,693,695,697,699,701,703,705,707,709,711,713,715,717,719,721,723,725,727,729,731,733,735,737,739,741,743,745,747,749,751,753,755,757,759,761,763,765,767,769,771,773,775,777,779,781,783,785,787,789,791,793,795,797,799,801,803,805,807,809,811,813,815,817,819,821,823,825,827,829,831,833,835,837,839,841,843,845,847,849,851,853,855,857,859,861,863,865,867,869,871,873,875,877,879,881,883,885,887,889,891,893,895,897,899,901,903,905,907,909,911,913,915,917,919,921,923,925,927,929,931,933,935,937,939,941,943,945,947,949,951,953,955,957,959,961,963,965,967,969,971,973,975,977,979,981,983,985,987,989,991,993,995,997,999,1001,1003,1005,1007,1009,1011,1013,1015,1017,1019,1021,1023,1025,1027,1029,1031,1033,1035,1037,1039,1041,1043,1045,1047,1049,1051,1053,1055,1057,1059,1061,1063,1065,1067,1069,1071,1073,1075,1077,1079,1081,1083,1085,1087,1089,1091,1093,1095,1097,1099,1101,1103,1105,1107,1109,1111,1113,1115,1117,1119,1121,1123,1125,1127,1129,1131,1133,1135,1137,1139,1141,1143,1145,1147,1149,1151,1153,1155,1157,1159,1161,1163,1165,1167,1169,1171,1173,1175,1177,1179,1181,1183,1185,1187,1189,1191,1193,1195,1197,1199,1201,1203,1205,1207,1209,1211,1213,1215,1217,1219,1221,1223,1225,1227,1229,1231,1233,1235,1237,1239,1241,1243,1245,1247,1249,1251,1253,1255,1257,1259,1261,1263,1265,1267,1269,1271,1273,1275,1277,1279,1281,1283,1285,1287,1289,1291,1293,1295,1297,1299,1301,1303,1305,1307,1309,1311,1313,1315,1317,1319,1321,1323,1325,1327,1329,1331,1333,1335,1337,1339,1341,1343,1345,1347,1349,1351,1353,1355,1357,1359,1361,1363,1365,1367,1369,1371,1373,1375,1377,1379,1381,1383,1385,1387,1389,1391,1393,1395,1397,1399,1401,1403,1405,1407,1409,1411,1413,1415,1417,1419,1421,1423,1425,1427,1429,1431,1433,1435,1437,1439,1441,1443,1445,1447,1449,1451,1453,1455,1457,1459,1461,1463,1465,1467,1469,1471,1473,1475,1477,1479,1481,1483,1485,1487,1489,1491,1493,1495,1497,1499,1501,1503,1505,1507,1509,1511,1513,1515,1517,1519,1521,1523,1525,1527,1529,1531,1533,1535,1537,1539,1541,1543,1545,1547,1549,1551,1553,1555,1557,1559,1561,1563,1565,1567,1569,1571,1573,1575,1577,1579,1581,1583,1585,1587,1589,1591,1593,1595,1597,1599,1601,1603,1605,1607,1609,1611,1613,1615,1617,1619,1621,1623,1625,1627,1629,1631,1633,1635,1637,1639,1641,1643,1645,1647,1649,1651,1653,1655,1657,1659,1661,1663,1665,1667,1669,1671,1673,1675,1677,1679,1681,1683,1685,1687,1689,1691,1693,1695,1697,1699,1701,1703,1705,1707,1709,1711,1713,1715,1717,1719,1721,1723,1725,1727,1729,1731,1733,1735,1737,1739,1741,1743,1745,1747,1749,1751,1753,1755,1757,1759,1761,1763,1765,1767,1769,1771,1773,1775,1777,1779,1781,1783,1785,1787,1789,1791,1793,1795,1797,1799,1801,1803,1805,1807,1809,1811,1813,1815,1817,1819,1821,1823,1825,1827,1829,1831,1833,1835,1837,1839,1841,1843,1845,1847,1849,1851,1853,1855,1857,1859,1861,1863,1865,1867,1869,1871,1873,1875,1877,1879,1881,1883,1885,1887,1889,1891,1893,1895,1897,1899,1901,1903,1905,1907,1909,1911,1913,1915,1917,1919,1921,1923,1925,1927,1929,1931,1933,1935,1937,1939,1941,1943,1945,1947,1949,1951,1953,1955,1957,1959,1961,1963,1965,1967,1969,1971,1973,1975,1977,1979,1981,1983,1985,1987,1989,1991,1993,1995,1997,1999,2001,2003,2005,2007,2009,2011,2013,2015,2017,2019,2021,2023,2025,2027,2029,2031,2033,2035,2037,2039,2041,2043,2045,2047,2049,2051,2053,2055,2057,2059,2061,2063,2065,2067,2069,2071,2073,2075,2077,2079,2081,2083,2085,2087,2089,2091,2093,2095,2097,2099,2101,2103,2105,2107,2109,2111,2113,2115,2117,2119,2121,2123,2125,2127,2129,2131,2133,2135,2137,2139,2141,2143,2145,2147,2149,2151,2153,2155,2157,2159,2161,2163,2165,2167,2169,2171,2173,2175,2177,2179,2181,2183,2185,2187,2189,2191,2193,2195,2197,2199,2201,2203,2205,2207,2209,2211,2213,2215,2217,2219,2221,2223,2225,2227,2229,2231,2233,2235,2237,2239,2241,2243,2245,2247,2249,2251,2253,2255,2257,2259,2261,2263,2265,2267,2269,2271,2273,2275,2277,2279,2281,2283,2285,2287,2289,2291,2293,2295,2297,2299,2301,2303,2305,2307,2309,2311,2313,2315,2317,2319,2321,2323,2325,2327,2329,2331,2333,2335,2337,2339,2341,2343,2345,2347,2349,2351,2353,2355,2357,2359,2361,2363,2365,2367,2369,2371,2373,2375,2377,2379,2381,2383,2385,2387,2389,2391,2393,2395,2397,2399,2401,2403,2405,2407,2409,2411,2413,2415,2417,2419,2421,2423,2425,2427,2429,2431,2433,2435,2437,2439,2441,2443,2445,2447,2449,2451,2453,2455,2457,2459,2461,2463,2465,2467,2469,2471,2473,2475,2477,2479,2481,2483,2485,2487,2489,2491,2493,2495,2497,2499,2501,2503,2505,2507,2509,2511,2513,2515,2517,2519,2521,2523,2525,2527,2529,2531,2533,2535,2537,2539,2541,2543,2545,2547,2549,2551,2553,2555,2557,2559,2561,2563,2565,2567,2569,2571,2573,2575,2577,2579,2581,2583,2585,2587,2589,2591,2593,2595,2597,2599,2601,2603,2605,2607,2609,2611,2613,2615,2617,2619,2621,2623,2625,2627,2629,2631,2633,2635,2637,2639,2641,2643,2645,2647,2649,2651,2653,2655,2657,2659,2661,2663,2665,2667,2669,2671,2673,2675,2677,2679,2681,2683,2685,2687,2689,2691,2693,2695,2697,2699,2701,2703,2705,2707,2709,2711,2713,2715,2717,2719,2721,2723,2725,2727,2729,2731,2733,2735,2737,2739,2741,2743,2745,2747,2749,2751,2753,2755,2757,2759,2761,2763,2765,2767,2769,2771,2773,2775,2777,2779,2781,2783,2785,2787,2789,2791,2793,2795,2797,2799,2801,2803,2805,2807,2809,2811,2813,2815,2817,2819,2821,2823,2825,2827,2829,2831,2833,2835,2837,2839,2841,2843,2845,2847,2849,2851,2853,2855,2857,2859,2861,2863,2865,2867,2869,2871,2873,2875,2877,2879,2881,2883,2885,2887,2889,2891,2893,2895,2897,2899,2901,2903,2905,2907,2909,2911,2913,2915,2917,2919,2921,2923,2925,2927,2929,2931,2933,2935,2937,2939,2941,2943,2945,2947,2949,2951,2953,2955,2957,2959,2961,2963,2965,2967,2969,2971,2973,2975,2977,2979,2981,2983,2985,2987,2989,2991,2993,2995,2997,2999,3001,3003,3005,3007,3009,3011,3013,3015,3017,3019,3021,3023,3025,3027,3029,3031,3033,3035,3037,3039,3041,3043,3045,3047,3049,3051,3053,3055,3057,3059,3061,3063,3065,3067,3069,3071,3073,3075,3077,3079,3081,3083,3085,3087,3089,3091,3093,3095,3097,3099,3101,3103,3105,3107,3109,3111,3113,3115,3117,3119,3121,3123,3125,3127,3129,3131,3133,3135,3137,3139,3141,3143,3145,3147,3149,3151,3153,3155,3157,3159,3161,3163,3165,3167,3169,3171,3173,3175,3177,3179,3181,3183,3185,3187,3189,3191,3193,3195,3197,3199,3201,3203,3205,3207,3209,3211,3213,3215,3217,3219,3221,3223,3225,3227,3229,3231,3233,3235,3237,3239,3241,3243,3245,3247,3249,3251,3253,3255,3257,3259,3261,3263,3265,3267,3269,3271,3273,3275,3277,3279,3281,3283,3285,3287,3289,3291,3293,3295,3297,3299,3301,3303,3305,3307,3309,3311,3313,3315,3317,3319,3321,3323,3325,3327,3329,3331,3333,3335,3337,3339,3341,3343,3345,3347,3349,3351,3353,3355,3357,3359,3361,3363,3365,3367,3369,3371,3373,3375,3377,3379,3381,3383,3385,3387,3389,3391,3393,3395,3397,3399,3401,3403,3405,3407,3409,3411,3413,3415,3417,3419,3421,3423,3425,3427,3429,3431,3433,3435,3437,3439,3441,3443,3445,3447,3449,3451,3453,3455,3457,3459,3461,3463,3465,3467,3469,3471,3473,3475,3477,3479,3481,3483,3485,3487,3489,3491,3493,3495,3497,3499,3501,3503,3505,3507,3509,3511,3513,3515,3517,3519,3521,3523,3525,3527,3529,3531,3533,3535,3537,3539,3541,3543,3545,3547,3549,3551,3553,3555,3557,3559,3561,3563,3565,3567,3569,3571,3573,3575,3577,3579,3581,3583,3585,3587,3589,3591,3593,3595,3597,3599,3601,3603,3605,3607,3609,3611,3613,3615,3617,3619,3621,3623,3625,3627,3629,3631,3633,3635,3637,3639,3641,3643,3645,3647,3649,3651,3653],{"categories":88},[89],"Developer Productivity",{"categories":91},[92],"Business & SaaS",{"categories":94},[47],{"categories":96},[97],"AI Automation",{"categories":99},[100],"Product Strategy",{"categories":102},[47],{"categories":104},[89],{"categories":106},[92],{"categories":108},[],{"categories":110},[47],{"categories":112},[],{"categories":114},[115],"AI News & Trends",{"categories":117},[97],{"categories":119},[115],{"categories":121},[97],{"categories":123},[97],{"categories":125},[47],{"categories":127},[47],{"categories":129},[115],{"categories":131},[47],{"categories":133},[],{"categories":135},[136],"Design & Frontend",{"categories":138},[139],"Data Science & Visualization",{"categories":141},[115],{"categories":143},[],{"categories":145},[146],"Software Engineering",{"categories":148},[47],{"categories":150},[97],{"categories":152},[153],"Marketing & Growth",{"categories":155},[47],{"categories":157},[97],{"categories":159},[],{"categories":161},[],{"categories":163},[136],{"categories":165},[97],{"categories":167},[89],{"categories":169},[136],{"categories":171},[47],{"categories":173},[97],{"categories":175},[115],{"categories":177},[],{"categories":179},[],{"categories":181},[97],{"categories":183},[146],{"categories":185},[],{"categories":187},[92],{"categories":189},[],{"categories":191},[],{"categories":193},[97],{"categories":195},[97],{"categories":197},[47],{"categories":199},[],{"categories":201},[146],{"categories":203},[],{"categories":205},[],{"categories":207},[],{"categories":209},[47],{"categories":211},[153],{"categories":213},[136],{"categories":215},[136],{"categories":217},[47],{"categories":219},[97],{"categories":221},[47],{"categories":223},[47],{"categories":225},[97],{"categories":227},[97],{"categories":229},[139],{"categories":231},[115],{"categories":233},[97],{"categories":235},[153],{"categories":237},[97],{"categories":239},[100],{"categories":241},[],{"categories":243},[97],{"categories":245},[],{"categories":247},[97],{"categories":249},[146],{"categories":251},[136],{"categories":253},[47],{"categories":255},[],{"categories":257},[],{"categories":259},[97],{"categories":261},[],{"categories":263},[47],{"categories":265},[],{"categories":267},[89],{"categories":269},[146],{"categories":271},[92],{"categories":273},[115],{"categories":275},[47],{"categories":277},[],{"categories":279},[47],{"categories":281},[],{"categories":283},[146],{"categories":285},[139],{"categories":287},[],{"categories":289},[47],{"categories":291},[136],{"categories":293},[],{"categories":295},[136],{"categories":297},[97],{"categories":299},[],{"categories":301},[97],{"categories":303},[115],{"categories":305},[47],{"categories":307},[],{"categories":309},[97],{"categories":311},[47],{"categories":313},[100],{"categories":315},[],{"categories":317},[47],{"categories":319},[97],{"categories":321},[97],{"categories":323},[],{"categories":325},[139],{"categories":327},[47],{"categories":329},[],{"categories":331},[89],{"categories":333},[92],{"categories":335},[47],{"categories":337},[97],{"categories":339},[146],{"categories":341},[47],{"categories":343},[],{"categories":345},[],{"categories":347},[47],{"categories":349},[],{"categories":351},[136],{"categories":353},[],{"categories":355},[47],{"categories":357},[],{"categories":359},[97],{"categories":361},[47],{"categories":363},[136],{"categories":365},[],{"categories":367},[47],{"categories":369},[47],{"categories":371},[92],{"categories":373},[97],{"categories":375},[47],{"categories":377},[136],{"categories":379},[97],{"categories":381},[],{"categories":383},[],{"categories":385},[115],{"categories":387},[],{"categories":389},[47],{"categories":391},[92,153],{"categories":393},[],{"categories":395},[47],{"categories":397},[],{"categories":399},[],{"categories":401},[47],{"categories":403},[],{"categories":405},[47],{"categories":407},[408],"DevOps & Cloud",{"categories":410},[],{"categories":412},[115],{"categories":414},[136],{"categories":416},[],{"categories":418},[115],{"categories":420},[115],{"categories":422},[47],{"categories":424},[153],{"categories":426},[],{"categories":428},[92],{"categories":430},[],{"categories":432},[47,408],{"categories":434},[47],{"categories":436},[47],{"categories":438},[97],{"categories":440},[47,146],{"categories":442},[139],{"categories":444},[47],{"categories":446},[153],{"categories":448},[97],{"categories":450},[97],{"categories":452},[],{"categories":454},[97],{"categories":456},[47,92],{"categories":458},[],{"categories":460},[136],{"categories":462},[136],{"categories":464},[],{"categories":466},[],{"categories":468},[115],{"categories":470},[],{"categories":472},[89],{"categories":474},[146],{"categories":476},[47],{"categories":478},[136],{"categories":480},[97],{"categories":482},[146],{"categories":484},[115],{"categories":486},[136],{"categories":488},[],{"categories":490},[47],{"categories":492},[47],{"categories":494},[47],{"categories":496},[115],{"categories":498},[89],{"categories":500},[47],{"categories":502},[97],{"categories":504},[408],{"categories":506},[136],{"categories":508},[97],{"categories":510},[],{"categories":512},[],{"categories":514},[136],{"categories":516},[115],{"categories":518},[139],{"categories":520},[],{"categories":522},[47],{"categories":524},[47],{"categories":526},[92],{"categories":528},[47],{"categories":530},[47],{"categories":532},[115],{"categories":534},[],{"categories":536},[97],{"categories":538},[146],{"categories":540},[],{"categories":542},[47],{"categories":544},[47],{"categories":546},[97],{"categories":548},[],{"categories":550},[],{"categories":552},[47],{"categories":554},[],{"categories":556},[92],{"categories":558},[97],{"categories":560},[],{"categories":562},[89],{"categories":564},[47],{"categories":566},[92],{"categories":568},[115],{"categories":570},[],{"categories":572},[],{"categories":574},[],{"categories":576},[115],{"categories":578},[115],{"categories":580},[],{"categories":582},[],{"categories":584},[92],{"categories":586},[],{"categories":588},[],{"categories":590},[89],{"categories":592},[],{"categories":594},[153],{"categories":596},[97],{"categories":598},[92],{"categories":600},[97],{"categories":602},[],{"categories":604},[100],{"categories":606},[136],{"categories":608},[146],{"categories":610},[47],{"categories":612},[97],{"categories":614},[92],{"categories":616},[47],{"categories":618},[],{"categories":620},[],{"categories":622},[146],{"categories":624},[139],{"categories":626},[100],{"categories":628},[97],{"categories":630},[47],{"categories":632},[],{"categories":634},[408],{"categories":636},[],{"categories":638},[97],{"categories":640},[],{"categories":642},[],{"categories":644},[47],{"categories":646},[136],{"categories":648},[153],{"categories":650},[97],{"categories":652},[],{"categories":654},[89],{"categories":656},[],{"categories":658},[115],{"categories":660},[47,408],{"categories":662},[115],{"categories":664},[47],{"categories":666},[92],{"categories":668},[47],{"categories":670},[],{"categories":672},[92],{"categories":674},[],{"categories":676},[146],{"categories":678},[136],{"categories":680},[115],{"categories":682},[139],{"categories":684},[89],{"categories":686},[47],{"categories":688},[146],{"categories":690},[],{"categories":692},[],{"categories":694},[100],{"categories":696},[],{"categories":698},[47],{"categories":700},[],{"categories":702},[136],{"categories":704},[136],{"categories":706},[136],{"categories":708},[],{"categories":710},[],{"categories":712},[115],{"categories":714},[97],{"categories":716},[47],{"categories":718},[47],{"categories":720},[47],{"categories":722},[92],{"categories":724},[47],{"categories":726},[],{"categories":728},[146],{"categories":730},[146],{"categories":732},[92],{"categories":734},[],{"categories":736},[47],{"categories":738},[47],{"categories":740},[92],{"categories":742},[115],{"categories":744},[153],{"categories":746},[97],{"categories":748},[],{"categories":750},[136],{"categories":752},[],{"categories":754},[47],{"categories":756},[],{"categories":758},[92],{"categories":760},[97],{"categories":762},[],{"categories":764},[408],{"categories":766},[139],{"categories":768},[146],{"categories":770},[153],{"categories":772},[146],{"categories":774},[97],{"categories":776},[],{"categories":778},[],{"categories":780},[97],{"categories":782},[89],{"categories":784},[97],{"categories":786},[100],{"categories":788},[92],{"categories":790},[],{"categories":792},[47],{"categories":794},[100],{"categories":796},[47],{"categories":798},[47],{"categories":800},[153],{"categories":802},[136],{"categories":804},[97],{"categories":806},[],{"categories":808},[],{"categories":810},[408],{"categories":812},[146],{"categories":814},[],{"categories":816},[97],{"categories":818},[47],{"categories":820},[136,47],{"categories":822},[89],{"categories":824},[],{"categories":826},[47],{"categories":828},[89],{"categories":830},[136],{"categories":832},[97],{"categories":834},[146],{"categories":836},[],{"categories":838},[47],{"categories":840},[],{"categories":842},[89],{"categories":844},[],{"categories":846},[97],{"categories":848},[100],{"categories":850},[47],{"categories":852},[47],{"categories":854},[136],{"categories":856},[97],{"categories":858},[408],{"categories":860},[136],{"categories":862},[97],{"categories":864},[47],{"categories":866},[47],{"categories":868},[47],{"categories":870},[115],{"categories":872},[],{"categories":874},[100],{"categories":876},[97],{"categories":878},[136],{"categories":880},[97],{"categories":882},[146],{"categories":884},[136],{"categories":886},[97],{"categories":888},[115],{"categories":890},[],{"categories":892},[47],{"categories":894},[136],{"categories":896},[47],{"categories":898},[89],{"categories":900},[115],{"categories":902},[47],{"categories":904},[153],{"categories":906},[47],{"categories":908},[47],{"categories":910},[97],{"categories":912},[97],{"categories":914},[47],{"categories":916},[97],{"categories":918},[136],{"categories":920},[47],{"categories":922},[],{"categories":924},[],{"categories":926},[146],{"categories":928},[],{"categories":930},[89],{"categories":932},[408],{"categories":934},[],{"categories":936},[89],{"categories":938},[92],{"categories":940},[153],{"categories":942},[],{"categories":944},[92],{"categories":946},[],{"categories":948},[],{"categories":950},[],{"categories":952},[],{"categories":954},[],{"categories":956},[47],{"categories":958},[97],{"categories":960},[408],{"categories":962},[89],{"categories":964},[47],{"categories":966},[146],{"categories":968},[100],{"categories":970},[47],{"categories":972},[153],{"categories":974},[47],{"categories":976},[47],{"categories":978},[47],{"categories":980},[47,89],{"categories":982},[146],{"categories":984},[146],{"categories":986},[136],{"categories":988},[47],{"categories":990},[],{"categories":992},[],{"categories":994},[],{"categories":996},[146],{"categories":998},[139],{"categories":1000},[115],{"categories":1002},[136],{"categories":1004},[],{"categories":1006},[47],{"categories":1008},[47],{"categories":1010},[],{"categories":1012},[],{"categories":1014},[97],{"categories":1016},[47],{"categories":1018},[92],{"categories":1020},[],{"categories":1022},[89],{"categories":1024},[47],{"categories":1026},[89],{"categories":1028},[47],{"categories":1030},[146],{"categories":1032},[153],{"categories":1034},[47,136],{"categories":1036},[115],{"categories":1038},[136],{"categories":1040},[],{"categories":1042},[408],{"categories":1044},[136],{"categories":1046},[97],{"categories":1048},[],{"categories":1050},[],{"categories":1052},[],{"categories":1054},[],{"categories":1056},[146],{"categories":1058},[97],{"categories":1060},[97],{"categories":1062},[47],{"categories":1064},[47],{"categories":1066},[],{"categories":1068},[136],{"categories":1070},[],{"categories":1072},[],{"categories":1074},[97],{"categories":1076},[],{"categories":1078},[],{"categories":1080},[153],{"categories":1082},[153],{"categories":1084},[97],{"categories":1086},[],{"categories":1088},[47],{"categories":1090},[47],{"categories":1092},[146],{"categories":1094},[136],{"categories":1096},[136],{"categories":1098},[97],{"categories":1100},[89],{"categories":1102},[47],{"categories":1104},[136],{"categories":1106},[136],{"categories":1108},[97],{"categories":1110},[97],{"categories":1112},[47],{"categories":1114},[],{"categories":1116},[],{"categories":1118},[47],{"categories":1120},[97],{"categories":1122},[115],{"categories":1124},[146],{"categories":1126},[89],{"categories":1128},[47],{"categories":1130},[],{"categories":1132},[97],{"categories":1134},[97],{"categories":1136},[],{"categories":1138},[89],{"categories":1140},[47],{"categories":1142},[89],{"categories":1144},[89],{"categories":1146},[],{"categories":1148},[],{"categories":1150},[97],{"categories":1152},[97],{"categories":1154},[47],{"categories":1156},[47],{"categories":1158},[115],{"categories":1160},[139],{"categories":1162},[100],{"categories":1164},[115],{"categories":1166},[136],{"categories":1168},[],{"categories":1170},[115],{"categories":1172},[],{"categories":1174},[],{"categories":1176},[],{"categories":1178},[],{"categories":1180},[146],{"categories":1182},[139],{"categories":1184},[],{"categories":1186},[47],{"categories":1188},[47],{"categories":1190},[139],{"categories":1192},[146],{"categories":1194},[],{"categories":1196},[],{"categories":1198},[97],{"categories":1200},[115],{"categories":1202},[115],{"categories":1204},[97],{"categories":1206},[89],{"categories":1208},[47,408],{"categories":1210},[],{"categories":1212},[136],{"categories":1214},[89],{"categories":1216},[97],{"categories":1218},[136],{"categories":1220},[],{"categories":1222},[97],{"categories":1224},[97],{"categories":1226},[47],{"categories":1228},[153],{"categories":1230},[146],{"categories":1232},[136],{"categories":1234},[],{"categories":1236},[97],{"categories":1238},[47],{"categories":1240},[97],{"categories":1242},[97],{"categories":1244},[97],{"categories":1246},[153],{"categories":1248},[97],{"categories":1250},[47],{"categories":1252},[],{"categories":1254},[153],{"categories":1256},[115],{"categories":1258},[97],{"categories":1260},[],{"categories":1262},[],{"categories":1264},[47],{"categories":1266},[97],{"categories":1268},[115],{"categories":1270},[97],{"categories":1272},[],{"categories":1274},[],{"categories":1276},[],{"categories":1278},[97],{"categories":1280},[],{"categories":1282},[],{"categories":1284},[139],{"categories":1286},[47],{"categories":1288},[139],{"categories":1290},[115],{"categories":1292},[47],{"categories":1294},[47],{"categories":1296},[97],{"categories":1298},[47],{"categories":1300},[],{"categories":1302},[],{"categories":1304},[408],{"categories":1306},[],{"categories":1308},[],{"categories":1310},[89],{"categories":1312},[],{"categories":1314},[],{"categories":1316},[],{"categories":1318},[],{"categories":1320},[146],{"categories":1322},[115],{"categories":1324},[153],{"categories":1326},[92],{"categories":1328},[47],{"categories":1330},[47],{"categories":1332},[92],{"categories":1334},[],{"categories":1336},[136],{"categories":1338},[97],{"categories":1340},[92],{"categories":1342},[47],{"categories":1344},[47],{"categories":1346},[89],{"categories":1348},[],{"categories":1350},[89],{"categories":1352},[47],{"categories":1354},[153],{"categories":1356},[97],{"categories":1358},[115],{"categories":1360},[92],{"categories":1362},[47],{"categories":1364},[97],{"categories":1366},[],{"categories":1368},[47],{"categories":1370},[89],{"categories":1372},[47],{"categories":1374},[],{"categories":1376},[115],{"categories":1378},[47],{"categories":1380},[],{"categories":1382},[92],{"categories":1384},[47],{"categories":1386},[],{"categories":1388},[],{"categories":1390},[],{"categories":1392},[47],{"categories":1394},[],{"categories":1396},[408],{"categories":1398},[47],{"categories":1400},[],{"categories":1402},[47],{"categories":1404},[47],{"categories":1406},[47],{"categories":1408},[47,408],{"categories":1410},[47],{"categories":1412},[47],{"categories":1414},[136],{"categories":1416},[97],{"categories":1418},[],{"categories":1420},[97],{"categories":1422},[47],{"categories":1424},[47],{"categories":1426},[47],{"categories":1428},[89],{"categories":1430},[89],{"categories":1432},[146],{"categories":1434},[136],{"categories":1436},[97],{"categories":1438},[],{"categories":1440},[47],{"categories":1442},[115],{"categories":1444},[47],{"categories":1446},[92],{"categories":1448},[],{"categories":1450},[408],{"categories":1452},[136],{"categories":1454},[136],{"categories":1456},[97],{"categories":1458},[115],{"categories":1460},[97],{"categories":1462},[47],{"categories":1464},[],{"categories":1466},[47],{"categories":1468},[],{"categories":1470},[],{"categories":1472},[47],{"categories":1474},[47],{"categories":1476},[47],{"categories":1478},[97],{"categories":1480},[47],{"categories":1482},[],{"categories":1484},[139],{"categories":1486},[97],{"categories":1488},[],{"categories":1490},[47],{"categories":1492},[115],{"categories":1494},[],{"categories":1496},[136],{"categories":1498},[408],{"categories":1500},[115],{"categories":1502},[146],{"categories":1504},[146],{"categories":1506},[115],{"categories":1508},[115],{"categories":1510},[408],{"categories":1512},[],{"categories":1514},[115],{"categories":1516},[47],{"categories":1518},[89],{"categories":1520},[115],{"categories":1522},[],{"categories":1524},[139],{"categories":1526},[115],{"categories":1528},[146],{"categories":1530},[115],{"categories":1532},[408],{"categories":1534},[47],{"categories":1536},[47],{"categories":1538},[],{"categories":1540},[92],{"categories":1542},[],{"categories":1544},[],{"categories":1546},[47],{"categories":1548},[47],{"categories":1550},[47],{"categories":1552},[47],{"categories":1554},[],{"categories":1556},[139],{"categories":1558},[89],{"categories":1560},[],{"categories":1562},[47],{"categories":1564},[47],{"categories":1566},[408],{"categories":1568},[408],{"categories":1570},[],{"categories":1572},[97],{"categories":1574},[115],{"categories":1576},[115],{"categories":1578},[47],{"categories":1580},[97],{"categories":1582},[],{"categories":1584},[136],{"categories":1586},[47],{"categories":1588},[47],{"categories":1590},[],{"categories":1592},[],{"categories":1594},[408],{"categories":1596},[47],{"categories":1598},[146],{"categories":1600},[92],{"categories":1602},[47],{"categories":1604},[],{"categories":1606},[97],{"categories":1608},[89],{"categories":1610},[89],{"categories":1612},[],{"categories":1614},[47],{"categories":1616},[136],{"categories":1618},[97],{"categories":1620},[],{"categories":1622},[47],{"categories":1624},[47],{"categories":1626},[97],{"categories":1628},[],{"categories":1630},[97],{"categories":1632},[146],{"categories":1634},[],{"categories":1636},[47],{"categories":1638},[],{"categories":1640},[47],{"categories":1642},[],{"categories":1644},[47],{"categories":1646},[47],{"categories":1648},[],{"categories":1650},[47],{"categories":1652},[115],{"categories":1654},[47],{"categories":1656},[47],{"categories":1658},[89],{"categories":1660},[47],{"categories":1662},[115],{"categories":1664},[97],{"categories":1666},[],{"categories":1668},[47],{"categories":1670},[153],{"categories":1672},[],{"categories":1674},[],{"categories":1676},[],{"categories":1678},[89],{"categories":1680},[115],{"categories":1682},[97],{"categories":1684},[47],{"categories":1686},[136],{"categories":1688},[97],{"categories":1690},[],{"categories":1692},[97],{"categories":1694},[],{"categories":1696},[47],{"categories":1698},[97],{"categories":1700},[47],{"categories":1702},[],{"categories":1704},[47],{"categories":1706},[47],{"categories":1708},[115],{"categories":1710},[136],{"categories":1712},[97],{"categories":1714},[136],{"categories":1716},[92],{"categories":1718},[],{"categories":1720},[],{"categories":1722},[47],{"categories":1724},[89],{"categories":1726},[115],{"categories":1728},[],{"categories":1730},[],{"categories":1732},[146],{"categories":1734},[136],{"categories":1736},[],{"categories":1738},[47],{"categories":1740},[],{"categories":1742},[153],{"categories":1744},[47],{"categories":1746},[408],{"categories":1748},[146],{"categories":1750},[],{"categories":1752},[97],{"categories":1754},[47],{"categories":1756},[97],{"categories":1758},[97],{"categories":1760},[47],{"categories":1762},[],{"categories":1764},[89],{"categories":1766},[47],{"categories":1768},[92],{"categories":1770},[146],{"categories":1772},[136],{"categories":1774},[],{"categories":1776},[],{"categories":1778},[],{"categories":1780},[97],{"categories":1782},[136],{"categories":1784},[115],{"categories":1786},[47],{"categories":1788},[115],{"categories":1790},[136],{"categories":1792},[],{"categories":1794},[136],{"categories":1796},[115],{"categories":1798},[92],{"categories":1800},[47],{"categories":1802},[115],{"categories":1804},[153],{"categories":1806},[],{"categories":1808},[],{"categories":1810},[139],{"categories":1812},[47,146],{"categories":1814},[115],{"categories":1816},[47],{"categories":1818},[97],{"categories":1820},[97],{"categories":1822},[47],{"categories":1824},[],{"categories":1826},[146],{"categories":1828},[47],{"categories":1830},[139],{"categories":1832},[97],{"categories":1834},[153],{"categories":1836},[408],{"categories":1838},[],{"categories":1840},[89],{"categories":1842},[97],{"categories":1844},[97],{"categories":1846},[146],{"categories":1848},[47],{"categories":1850},[47],{"categories":1852},[],{"categories":1854},[],{"categories":1856},[],{"categories":1858},[408],{"categories":1860},[115],{"categories":1862},[47],{"categories":1864},[47],{"categories":1866},[47],{"categories":1868},[],{"categories":1870},[139],{"categories":1872},[92],{"categories":1874},[],{"categories":1876},[97],{"categories":1878},[408],{"categories":1880},[],{"categories":1882},[136],{"categories":1884},[136],{"categories":1886},[],{"categories":1888},[146],{"categories":1890},[136],{"categories":1892},[47],{"categories":1894},[],{"categories":1896},[115],{"categories":1898},[47],{"categories":1900},[136],{"categories":1902},[97],{"categories":1904},[115],{"categories":1906},[],{"categories":1908},[97],{"categories":1910},[136],{"categories":1912},[47],{"categories":1914},[],{"categories":1916},[47],{"categories":1918},[47],{"categories":1920},[408],{"categories":1922},[115],{"categories":1924},[139],{"categories":1926},[139],{"categories":1928},[],{"categories":1930},[],{"categories":1932},[],{"categories":1934},[97],{"categories":1936},[146],{"categories":1938},[146],{"categories":1940},[],{"categories":1942},[],{"categories":1944},[47],{"categories":1946},[],{"categories":1948},[97],{"categories":1950},[47],{"categories":1952},[],{"categories":1954},[47],{"categories":1956},[92],{"categories":1958},[47],{"categories":1960},[153],{"categories":1962},[97],{"categories":1964},[47],{"categories":1966},[146],{"categories":1968},[115],{"categories":1970},[97],{"categories":1972},[],{"categories":1974},[115],{"categories":1976},[97],{"categories":1978},[97],{"categories":1980},[],{"categories":1982},[92],{"categories":1984},[97],{"categories":1986},[],{"categories":1988},[47],{"categories":1990},[89],{"categories":1992},[115],{"categories":1994},[408],{"categories":1996},[97],{"categories":1998},[97],{"categories":2000},[89],{"categories":2002},[47],{"categories":2004},[],{"categories":2006},[],{"categories":2008},[136],{"categories":2010},[47,92],{"categories":2012},[],{"categories":2014},[89],{"categories":2016},[139],{"categories":2018},[47],{"categories":2020},[146],{"categories":2022},[47],{"categories":2024},[97],{"categories":2026},[47],{"categories":2028},[47],{"categories":2030},[115],{"categories":2032},[97],{"categories":2034},[],{"categories":2036},[],{"categories":2038},[97],{"categories":2040},[47],{"categories":2042},[408],{"categories":2044},[],{"categories":2046},[47],{"categories":2048},[97],{"categories":2050},[],{"categories":2052},[47],{"categories":2054},[153],{"categories":2056},[139],{"categories":2058},[97],{"categories":2060},[47],{"categories":2062},[408],{"categories":2064},[],{"categories":2066},[47],{"categories":2068},[153],{"categories":2070},[136],{"categories":2072},[47],{"categories":2074},[],{"categories":2076},[153],{"categories":2078},[115],{"categories":2080},[47],{"categories":2082},[47],{"categories":2084},[89],{"categories":2086},[],{"categories":2088},[],{"categories":2090},[136],{"categories":2092},[47],{"categories":2094},[139],{"categories":2096},[153],{"categories":2098},[153],{"categories":2100},[115],{"categories":2102},[],{"categories":2104},[],{"categories":2106},[47],{"categories":2108},[],{"categories":2110},[47,146],{"categories":2112},[115],{"categories":2114},[97],{"categories":2116},[146],{"categories":2118},[47],{"categories":2120},[89],{"categories":2122},[],{"categories":2124},[],{"categories":2126},[89],{"categories":2128},[153],{"categories":2130},[47],{"categories":2132},[],{"categories":2134},[136,47],{"categories":2136},[408],{"categories":2138},[89],{"categories":2140},[],{"categories":2142},[92],{"categories":2144},[92],{"categories":2146},[47],{"categories":2148},[146],{"categories":2150},[97],{"categories":2152},[115],{"categories":2154},[153],{"categories":2156},[136],{"categories":2158},[47],{"categories":2160},[47],{"categories":2162},[47],{"categories":2164},[89],{"categories":2166},[47],{"categories":2168},[97],{"categories":2170},[115],{"categories":2172},[],{"categories":2174},[],{"categories":2176},[139],{"categories":2178},[146],{"categories":2180},[47],{"categories":2182},[136],{"categories":2184},[139],{"categories":2186},[47],{"categories":2188},[47],{"categories":2190},[97],{"categories":2192},[97],{"categories":2194},[47,92],{"categories":2196},[],{"categories":2198},[136],{"categories":2200},[],{"categories":2202},[47],{"categories":2204},[115],{"categories":2206},[89],{"categories":2208},[89],{"categories":2210},[97],{"categories":2212},[47],{"categories":2214},[92],{"categories":2216},[146],{"categories":2218},[153],{"categories":2220},[],{"categories":2222},[115],{"categories":2224},[47],{"categories":2226},[47],{"categories":2228},[115],{"categories":2230},[146],{"categories":2232},[47],{"categories":2234},[97],{"categories":2236},[115],{"categories":2238},[47],{"categories":2240},[136],{"categories":2242},[47],{"categories":2244},[47],{"categories":2246},[408],{"categories":2248},[100],{"categories":2250},[97],{"categories":2252},[47],{"categories":2254},[115],{"categories":2256},[97],{"categories":2258},[153],{"categories":2260},[47],{"categories":2262},[],{"categories":2264},[47],{"categories":2266},[],{"categories":2268},[],{"categories":2270},[],{"categories":2272},[92],{"categories":2274},[47],{"categories":2276},[97],{"categories":2278},[115],{"categories":2280},[115],{"categories":2282},[115],{"categories":2284},[115],{"categories":2286},[],{"categories":2288},[89],{"categories":2290},[97],{"categories":2292},[115],{"categories":2294},[89],{"categories":2296},[97],{"categories":2298},[47],{"categories":2300},[47,97],{"categories":2302},[97],{"categories":2304},[408],{"categories":2306},[115],{"categories":2308},[115],{"categories":2310},[97],{"categories":2312},[47],{"categories":2314},[],{"categories":2316},[115],{"categories":2318},[153],{"categories":2320},[89],{"categories":2322},[47],{"categories":2324},[47],{"categories":2326},[],{"categories":2328},[146],{"categories":2330},[],{"categories":2332},[89],{"categories":2334},[97],{"categories":2336},[115],{"categories":2338},[47],{"categories":2340},[115],{"categories":2342},[89],{"categories":2344},[115],{"categories":2346},[115],{"categories":2348},[],{"categories":2350},[92],{"categories":2352},[97],{"categories":2354},[115],{"categories":2356},[115],{"categories":2358},[115],{"categories":2360},[115],{"categories":2362},[115],{"categories":2364},[115],{"categories":2366},[115],{"categories":2368},[115],{"categories":2370},[115],{"categories":2372},[115],{"categories":2374},[139],{"categories":2376},[89],{"categories":2378},[47],{"categories":2380},[47],{"categories":2382},[],{"categories":2384},[47,89],{"categories":2386},[],{"categories":2388},[97],{"categories":2390},[115],{"categories":2392},[97],{"categories":2394},[47],{"categories":2396},[47],{"categories":2398},[47],{"categories":2400},[47],{"categories":2402},[47],{"categories":2404},[97],{"categories":2406},[92],{"categories":2408},[136],{"categories":2410},[115],{"categories":2412},[47],{"categories":2414},[],{"categories":2416},[],{"categories":2418},[97],{"categories":2420},[136],{"categories":2422},[47],{"categories":2424},[],{"categories":2426},[],{"categories":2428},[153],{"categories":2430},[47],{"categories":2432},[],{"categories":2434},[],{"categories":2436},[89],{"categories":2438},[92],{"categories":2440},[47],{"categories":2442},[92],{"categories":2444},[136],{"categories":2446},[],{"categories":2448},[115],{"categories":2450},[],{"categories":2452},[136],{"categories":2454},[47],{"categories":2456},[153],{"categories":2458},[],{"categories":2460},[153],{"categories":2462},[],{"categories":2464},[],{"categories":2466},[97],{"categories":2468},[],{"categories":2470},[92],{"categories":2472},[89],{"categories":2474},[136],{"categories":2476},[146],{"categories":2478},[],{"categories":2480},[],{"categories":2482},[47],{"categories":2484},[89],{"categories":2486},[153],{"categories":2488},[],{"categories":2490},[97],{"categories":2492},[97],{"categories":2494},[115],{"categories":2496},[47],{"categories":2498},[97],{"categories":2500},[47],{"categories":2502},[97],{"categories":2504},[47],{"categories":2506},[100],{"categories":2508},[115],{"categories":2510},[],{"categories":2512},[153],{"categories":2514},[146],{"categories":2516},[97],{"categories":2518},[],{"categories":2520},[47],{"categories":2522},[97],{"categories":2524},[92],{"categories":2526},[89],{"categories":2528},[47],{"categories":2530},[136],{"categories":2532},[146],{"categories":2534},[146],{"categories":2536},[47],{"categories":2538},[139],{"categories":2540},[47],{"categories":2542},[97],{"categories":2544},[92],{"categories":2546},[97],{"categories":2548},[47],{"categories":2550},[47],{"categories":2552},[97],{"categories":2554},[115],{"categories":2556},[],{"categories":2558},[89],{"categories":2560},[47],{"categories":2562},[97],{"categories":2564},[47],{"categories":2566},[47],{"categories":2568},[],{"categories":2570},[136],{"categories":2572},[92],{"categories":2574},[115],{"categories":2576},[47],{"categories":2578},[47],{"categories":2580},[136],{"categories":2582},[153],{"categories":2584},[139],{"categories":2586},[47],{"categories":2588},[115],{"categories":2590},[47],{"categories":2592},[97],{"categories":2594},[408],{"categories":2596},[47],{"categories":2598},[97],{"categories":2600},[139],{"categories":2602},[],{"categories":2604},[97],{"categories":2606},[146],{"categories":2608},[136],{"categories":2610},[47],{"categories":2612},[89],{"categories":2614},[92],{"categories":2616},[146],{"categories":2618},[],{"categories":2620},[97],{"categories":2622},[47],{"categories":2624},[],{"categories":2626},[115],{"categories":2628},[],{"categories":2630},[115],{"categories":2632},[47],{"categories":2634},[97],{"categories":2636},[97],{"categories":2638},[97],{"categories":2640},[],{"categories":2642},[],{"categories":2644},[47],{"categories":2646},[47],{"categories":2648},[],{"categories":2650},[136],{"categories":2652},[97],{"categories":2654},[153],{"categories":2656},[89],{"categories":2658},[],{"categories":2660},[],{"categories":2662},[115],{"categories":2664},[146],{"categories":2666},[47],{"categories":2668},[47],{"categories":2670},[47],{"categories":2672},[146],{"categories":2674},[115],{"categories":2676},[136],{"categories":2678},[47],{"categories":2680},[47],{"categories":2682},[47],{"categories":2684},[115],{"categories":2686},[47],{"categories":2688},[115],{"categories":2690},[97],{"categories":2692},[97],{"categories":2694},[146],{"categories":2696},[97],{"categories":2698},[47],{"categories":2700},[146],{"categories":2702},[136],{"categories":2704},[],{"categories":2706},[97],{"categories":2708},[],{"categories":2710},[],{"categories":2712},[92],{"categories":2714},[47],{"categories":2716},[97],{"categories":2718},[89],{"categories":2720},[97],{"categories":2722},[153],{"categories":2724},[],{"categories":2726},[97],{"categories":2728},[],{"categories":2730},[89],{"categories":2732},[97],{"categories":2734},[],{"categories":2736},[97],{"categories":2738},[47],{"categories":2740},[115],{"categories":2742},[47],{"categories":2744},[97],{"categories":2746},[115],{"categories":2748},[97],{"categories":2750},[146],{"categories":2752},[136],{"categories":2754},[89],{"categories":2756},[],{"categories":2758},[97],{"categories":2760},[136],{"categories":2762},[115],{"categories":2764},[47],{"categories":2766},[136],{"categories":2768},[89],{"categories":2770},[],{"categories":2772},[97],{"categories":2774},[97],{"categories":2776},[47],{"categories":2778},[],{"categories":2780},[97],{"categories":2782},[100],{"categories":2784},[115],{"categories":2786},[97],{"categories":2788},[92],{"categories":2790},[],{"categories":2792},[47],{"categories":2794},[100],{"categories":2796},[47],{"categories":2798},[97],{"categories":2800},[115],{"categories":2802},[89],{"categories":2804},[408],{"categories":2806},[47],{"categories":2808},[47],{"categories":2810},[47],{"categories":2812},[115],{"categories":2814},[92],{"categories":2816},[47],{"categories":2818},[136],{"categories":2820},[115],{"categories":2822},[408],{"categories":2824},[47],{"categories":2826},[],{"categories":2828},[],{"categories":2830},[408],{"categories":2832},[139],{"categories":2834},[97],{"categories":2836},[97],{"categories":2838},[115],{"categories":2840},[47],{"categories":2842},[89],{"categories":2844},[136],{"categories":2846},[97],{"categories":2848},[47],{"categories":2850},[153],{"categories":2852},[47],{"categories":2854},[97],{"categories":2856},[],{"categories":2858},[47],{"categories":2860},[47],{"categories":2862},[115],{"categories":2864},[89],{"categories":2866},[],{"categories":2868},[47],{"categories":2870},[47],{"categories":2872},[146],{"categories":2874},[136],{"categories":2876},[47,97],{"categories":2878},[153,92],{"categories":2880},[47],{"categories":2882},[],{"categories":2884},[97],{"categories":2886},[],{"categories":2888},[146],{"categories":2890},[47],{"categories":2892},[115],{"categories":2894},[],{"categories":2896},[97],{"categories":2898},[],{"categories":2900},[97],{"categories":2902},[89],{"categories":2904},[97],{"categories":2906},[47],{"categories":2908},[408],{"categories":2910},[153],{"categories":2912},[92],{"categories":2914},[92],{"categories":2916},[89],{"categories":2918},[89],{"categories":2920},[47],{"categories":2922},[97],{"categories":2924},[47],{"categories":2926},[47],{"categories":2928},[89],{"categories":2930},[47],{"categories":2932},[153],{"categories":2934},[115],{"categories":2936},[47],{"categories":2938},[97],{"categories":2940},[47],{"categories":2942},[],{"categories":2944},[146],{"categories":2946},[],{"categories":2948},[97],{"categories":2950},[89],{"categories":2952},[],{"categories":2954},[408],{"categories":2956},[47],{"categories":2958},[],{"categories":2960},[115],{"categories":2962},[97],{"categories":2964},[146],{"categories":2966},[47],{"categories":2968},[97],{"categories":2970},[146],{"categories":2972},[97],{"categories":2974},[115],{"categories":2976},[89],{"categories":2978},[115],{"categories":2980},[146],{"categories":2982},[47],{"categories":2984},[136],{"categories":2986},[47],{"categories":2988},[47],{"categories":2990},[47],{"categories":2992},[47],{"categories":2994},[97],{"categories":2996},[47],{"categories":2998},[97],{"categories":3000},[47],{"categories":3002},[89],{"categories":3004},[47],{"categories":3006},[97],{"categories":3008},[136],{"categories":3010},[89],{"categories":3012},[97],{"categories":3014},[136],{"categories":3016},[],{"categories":3018},[47],{"categories":3020},[47],{"categories":3022},[146],{"categories":3024},[],{"categories":3026},[97],{"categories":3028},[153],{"categories":3030},[47],{"categories":3032},[115],{"categories":3034},[153],{"categories":3036},[97],{"categories":3038},[92],{"categories":3040},[92],{"categories":3042},[47],{"categories":3044},[89],{"categories":3046},[],{"categories":3048},[47],{"categories":3050},[],{"categories":3052},[89],{"categories":3054},[47],{"categories":3056},[97],{"categories":3058},[97],{"categories":3060},[],{"categories":3062},[146],{"categories":3064},[146],{"categories":3066},[153],{"categories":3068},[136],{"categories":3070},[],{"categories":3072},[47],{"categories":3074},[89],{"categories":3076},[47],{"categories":3078},[146],{"categories":3080},[89],{"categories":3082},[115],{"categories":3084},[115],{"categories":3086},[],{"categories":3088},[115],{"categories":3090},[97],{"categories":3092},[136],{"categories":3094},[139],{"categories":3096},[47],{"categories":3098},[],{"categories":3100},[115],{"categories":3102},[146],{"categories":3104},[92],{"categories":3106},[47],{"categories":3108},[89],{"categories":3110},[408],{"categories":3112},[89],{"categories":3114},[],{"categories":3116},[],{"categories":3118},[115],{"categories":3120},[],{"categories":3122},[97],{"categories":3124},[97],{"categories":3126},[97],{"categories":3128},[],{"categories":3130},[47],{"categories":3132},[],{"categories":3134},[115],{"categories":3136},[89],{"categories":3138},[136],{"categories":3140},[47],{"categories":3142},[115],{"categories":3144},[115],{"categories":3146},[],{"categories":3148},[115],{"categories":3150},[89],{"categories":3152},[47],{"categories":3154},[],{"categories":3156},[97],{"categories":3158},[97],{"categories":3160},[89],{"categories":3162},[],{"categories":3164},[],{"categories":3166},[],{"categories":3168},[136],{"categories":3170},[97],{"categories":3172},[47],{"categories":3174},[],{"categories":3176},[],{"categories":3178},[],{"categories":3180},[136],{"categories":3182},[],{"categories":3184},[89],{"categories":3186},[],{"categories":3188},[],{"categories":3190},[136],{"categories":3192},[47],{"categories":3194},[115],{"categories":3196},[],{"categories":3198},[153],{"categories":3200},[115],{"categories":3202},[153],{"categories":3204},[47],{"categories":3206},[],{"categories":3208},[],{"categories":3210},[97],{"categories":3212},[],{"categories":3214},[],{"categories":3216},[97],{"categories":3218},[47],{"categories":3220},[],{"categories":3222},[97],{"categories":3224},[115],{"categories":3226},[153],{"categories":3228},[139],{"categories":3230},[97],{"categories":3232},[97],{"categories":3234},[],{"categories":3236},[],{"categories":3238},[],{"categories":3240},[115],{"categories":3242},[],{"categories":3244},[],{"categories":3246},[136],{"categories":3248},[89],{"categories":3250},[],{"categories":3252},[92],{"categories":3254},[153],{"categories":3256},[47],{"categories":3258},[146],{"categories":3260},[89],{"categories":3262},[139],{"categories":3264},[92],{"categories":3266},[146],{"categories":3268},[],{"categories":3270},[],{"categories":3272},[97],{"categories":3274},[89],{"categories":3276},[136],{"categories":3278},[89],{"categories":3280},[97],{"categories":3282},[408],{"categories":3284},[97],{"categories":3286},[],{"categories":3288},[47],{"categories":3290},[115],{"categories":3292},[146],{"categories":3294},[],{"categories":3296},[136],{"categories":3298},[115],{"categories":3300},[89],{"categories":3302},[97],{"categories":3304},[47],{"categories":3306},[92],{"categories":3308},[97,408],{"categories":3310},[97],{"categories":3312},[146],{"categories":3314},[47],{"categories":3316},[139],{"categories":3318},[153],{"categories":3320},[97],{"categories":3322},[],{"categories":3324},[97],{"categories":3326},[47],{"categories":3328},[92],{"categories":3330},[],{"categories":3332},[],{"categories":3334},[47],{"categories":3336},[139],{"categories":3338},[47],{"categories":3340},[],{"categories":3342},[115],{"categories":3344},[],{"categories":3346},[115],{"categories":3348},[146],{"categories":3350},[97],{"categories":3352},[47],{"categories":3354},[153],{"categories":3356},[146],{"categories":3358},[],{"categories":3360},[115],{"categories":3362},[47],{"categories":3364},[],{"categories":3366},[47],{"categories":3368},[97],{"categories":3370},[47],{"categories":3372},[97],{"categories":3374},[47],{"categories":3376},[47],{"categories":3378},[47],{"categories":3380},[47],{"categories":3382},[92],{"categories":3384},[],{"categories":3386},[100],{"categories":3388},[115],{"categories":3390},[47],{"categories":3392},[],{"categories":3394},[146],{"categories":3396},[47],{"categories":3398},[47],{"categories":3400},[97],{"categories":3402},[115],{"categories":3404},[47],{"categories":3406},[47],{"categories":3408},[92],{"categories":3410},[97],{"categories":3412},[136],{"categories":3414},[],{"categories":3416},[139],{"categories":3418},[47],{"categories":3420},[],{"categories":3422},[115],{"categories":3424},[153],{"categories":3426},[],{"categories":3428},[],{"categories":3430},[115],{"categories":3432},[115],{"categories":3434},[153],{"categories":3436},[89],{"categories":3438},[97],{"categories":3440},[97],{"categories":3442},[47],{"categories":3444},[92],{"categories":3446},[],{"categories":3448},[],{"categories":3450},[115],{"categories":3452},[139],{"categories":3454},[146],{"categories":3456},[97],{"categories":3458},[136],{"categories":3460},[139],{"categories":3462},[139],{"categories":3464},[],{"categories":3466},[115],{"categories":3468},[47],{"categories":3470},[47],{"categories":3472},[146],{"categories":3474},[],{"categories":3476},[115],{"categories":3478},[115],{"categories":3480},[115],{"categories":3482},[],{"categories":3484},[97],{"categories":3486},[47],{"categories":3488},[],{"categories":3490},[89],{"categories":3492},[92],{"categories":3494},[],{"categories":3496},[47],{"categories":3498},[47],{"categories":3500},[],{"categories":3502},[146],{"categories":3504},[],{"categories":3506},[],{"categories":3508},[],{"categories":3510},[],{"categories":3512},[47],{"categories":3514},[115],{"categories":3516},[],{"categories":3518},[],{"categories":3520},[47],{"categories":3522},[47],{"categories":3524},[47],{"categories":3526},[139],{"categories":3528},[47],{"categories":3530},[139],{"categories":3532},[],{"categories":3534},[139],{"categories":3536},[139],{"categories":3538},[408],{"categories":3540},[97],{"categories":3542},[146],{"categories":3544},[],{"categories":3546},[],{"categories":3548},[139],{"categories":3550},[146],{"categories":3552},[146],{"categories":3554},[146],{"categories":3556},[],{"categories":3558},[89],{"categories":3560},[146],{"categories":3562},[146],{"categories":3564},[89],{"categories":3566},[146],{"categories":3568},[92],{"categories":3570},[146],{"categories":3572},[146],{"categories":3574},[146],{"categories":3576},[139],{"categories":3578},[115],{"categories":3580},[115],{"categories":3582},[47],{"categories":3584},[146],{"categories":3586},[139],{"categories":3588},[408],{"categories":3590},[139],{"categories":3592},[139],{"categories":3594},[139],{"categories":3596},[],{"categories":3598},[92],{"categories":3600},[],{"categories":3602},[408],{"categories":3604},[146],{"categories":3606},[146],{"categories":3608},[146],{"categories":3610},[97],{"categories":3612},[115,92],{"categories":3614},[139],{"categories":3616},[],{"categories":3618},[],{"categories":3620},[139],{"categories":3622},[],{"categories":3624},[139],{"categories":3626},[115],{"categories":3628},[97],{"categories":3630},[],{"categories":3632},[146],{"categories":3634},[47],{"categories":3636},[136],{"categories":3638},[],{"categories":3640},[47],{"categories":3642},[],{"categories":3644},[115],{"categories":3646},[89],{"categories":3648},[139],{"categories":3650},[],{"categories":3652},[146],{"categories":3654},[115],[3656,3718,3864,3997],{"id":3657,"title":3658,"ai":3659,"body":3664,"categories":3692,"created_at":48,"date_modified":48,"description":40,"extension":49,"faq":48,"featured":50,"kicker_label":48,"meta":3693,"navigation":70,"path":3703,"published_at":3704,"question":48,"scraped_at":3705,"seo":3706,"sitemap":3707,"source_id":3708,"source_name":3709,"source_type":77,"source_url":3710,"stem":3711,"tags":3712,"thumbnail_url":48,"tldr":3715,"tweet":48,"unknown_tags":3716,"__hash__":3717},"summaries\u002Fsummaries\u002Fa3ec373360dad952-scale-genai-to-billions-of-rows-in-bigquery-at-94--summary.md","Scale GenAI to Billions of Rows in BigQuery at 94% Less Cost",{"provider":7,"model":8,"input_tokens":3660,"output_tokens":3661,"processing_time_ms":3662,"cost_usd":3663},4687,1619,26747,0.0017244,{"type":14,"value":3665,"toc":3687},[3666,3670,3673,3677,3680,3684],[17,3667,3669],{"id":3668},"replace-per-row-llm-calls-with-distilled-models-for-massive-savings","Replace Per-Row LLM Calls with Distilled Models for Massive Savings",[22,3671,3672],{},"Standard BigQuery AI functions like AI.CLASSIFY and AI.IF send every row to an LLM, burning through tokens and time on datasets with millions of rows—e.g., product reviews, claims, or support tickets. Optimized mode fixes this by automatically distilling a task-specific lightweight model: BigQuery samples your data, sends only that subset to the LLM for labeling, generates embeddings, and trains the distilled model locally on BigQuery compute. This model then processes the remaining rows using semantic embeddings for LLM-quality classification, filtering, or rating without full LLM inference per row. Result: process billions of rows at BigQuery speeds with drastically reduced latency and costs, as savings compound with data volume.",[17,3674,3676],{"id":3675},"trigger-optimization-automatically-or-with-one-parameter","Trigger Optimization Automatically or with One Parameter",[22,3678,3679],{},"No code rewrites needed—optimized mode activates for supported functions when you supply embeddings as a parameter (e.g., add embeddings column to AI.CLASSIFY) or if BigQuery's autonomous embeddings exist in the table. It auto-detects them, samples data, distills, and optimizes inline. For image analysis on 34k self-driving car camera shots, adding embeddings dropped tokens from 55M+ to 3M (94% reduction) and runtime from 16min to 2min, with  vast majority of rows processed by the distilled model. On 50k driver voice commands using AI.IF to filter 'slow down' requests, auto-detection optimized most rows without changes, delivering filtered results fast and cheap.",[17,3681,3683],{"id":3682},"trade-offs-and-when-to-use","Trade-offs and When to Use",[22,3685,3686],{},"Distillation trades full LLM flexibility for speed\u002Fcost on repetitive tasks like classification—ideal for large-scale filtering where you don't need per-row creativity. Quality matches LLM on samples and generalizes via embeddings; check job info tab post-query for optimization stats (e.g., % rows optimized). Start by adding embeddings to existing AI queries; scales best on growing datasets where per-row LLM becomes prohibitive.",{"title":40,"searchDepth":41,"depth":41,"links":3688},[3689,3690,3691],{"id":3668,"depth":41,"text":3669},{"id":3675,"depth":41,"text":3676},{"id":3682,"depth":41,"text":3683},[47],{"content_references":3694,"triage":3701},[3695,3698],{"type":59,"title":3696,"url":3697,"context":64},"Documentation for Optimized Mode","https:\u002F\u002Fgoo.gle\u002Foptimize-ai-functions",{"type":59,"title":3699,"url":3700,"context":64},"Generative AI in BigQuery overview","https:\u002F\u002Fgoo.gle\u002Fbq-genai-overview",{"relevance":66,"novelty":67,"quality":67,"actionability":67,"composite":68,"reasoning":3702},"Category: AI & LLMs. The article provides a detailed explanation of how to optimize LLM usage in BigQuery, addressing a specific pain point of cost and efficiency for AI-powered product builders. It offers actionable steps for implementing distilled models, making it highly relevant and practical.","\u002Fsummaries\u002Fa3ec373360dad952-scale-genai-to-billions-of-rows-in-bigquery-at-94-summary","2026-05-04 17:53:30","2026-05-05 16:07:55",{"title":3658,"description":40},{"loc":3703},"9a60decd09d8b7c9","Google Cloud Tech","https:\u002F\u002Fwww.youtube.com\u002Fwatch?v=-QLXKr94X6Q","summaries\u002Fa3ec373360dad952-scale-genai-to-billions-of-rows-in-bigquery-at-94--summary",[3713,81,82,3714],"data-science","embeddings","BigQuery's optimized mode distills LLMs into lightweight models using embeddings, slashing token use by 94% (55M to 3M) and query time from 16min to 2min on 34k images or 50k voice commands, scaling to billions of rows.",[81,82,3714],"HkxLBPyyl4DNze72ncpNs3IhUyMxhPDK2Q9X7o58TRA",{"id":3719,"title":3720,"ai":3721,"body":3726,"categories":3844,"created_at":48,"date_modified":48,"description":40,"extension":49,"faq":48,"featured":50,"kicker_label":48,"meta":3845,"navigation":70,"path":3849,"published_at":3850,"question":48,"scraped_at":3851,"seo":3852,"sitemap":3853,"source_id":3854,"source_name":3855,"source_type":77,"source_url":3856,"stem":3857,"tags":3858,"thumbnail_url":48,"tldr":3861,"tweet":48,"unknown_tags":3862,"__hash__":3863},"summaries\u002Fsummaries\u002F7b130eb6998f566d-scaling-llm-inference-kv-cache-batching-spec-decod-summary.md","Scaling LLM Inference: KV Cache, Batching, Spec Decoding & Multi-LoRA",{"provider":7,"model":8,"input_tokens":3722,"output_tokens":3723,"processing_time_ms":3724,"cost_usd":3725},8446,2303,12138,0.0028182,{"type":14,"value":3727,"toc":3835},[3728,3732,3735,3738,3742,3745,3748,3751,3754,3758,3761,3764,3767,3771,3774,3777,3780,3784,3787,3790,3793,3797,3800,3804],[17,3729,3731],{"id":3730},"inferences-memory-bound-reality-trumps-training-throughput","Inference's Memory-Bound Reality Trumps Training Throughput",[22,3733,3734],{},"Training optimizes for FLOPS throughput via parallel forward passes, but inference splits into prefill (compute-bound, parallel prompt processing populating KV cache) and decode (memory-bound, sequential token generation reading full weights + growing KV cache per step). \"Decode is memory-bandwidth-bound. The speed at which tokens are generated is not determined by how fast the GPU can multiply, but by how fast it can read. This is the single most important fact about LLM inference.\" KV cache dominates memory—scaling with sequence length, batch size, heads, dims, layers—often exceeding model weights at scale, inverting training assumptions where weights ruled.",[22,3736,3737],{},"Low arithmetic intensity in decode stalls GPUs waiting on HBM bandwidth, making peak FLOPS irrelevant. Hardware choices prioritize HBM capacity\u002Fbandwidth over compute. All optimizations reduce data movement or accelerate it.",[17,3739,3741],{"id":3740},"kv-cache-innovations-unlock-2-4x-throughput","KV Cache Innovations Unlock 2-4x Throughput",[22,3743,3744],{},"Naive allocation reserved max-sequence blocks per request, causing 20-40% utilization from fragmentation: internal (over-reserve) and external (scattered frees). PagedAttention (vLLM's core) uses fixed-size non-contiguous pages allocated dynamically—5th token gets 5th page, freeing instantly on completion—hitting 96% utilization for 2-4x throughput vs. HuggingFace Transformers.",[22,3746,3747],{},"RadixAttention (SGLang) adds prefix-sharing via radix tree for multi-turn chats\u002Ffew-shot\u002Fagent workflows: shared prefixes computed\u002Fstored once, 75-95% cache hits, up to 6.4x throughput on prefix-heavy loads with LRU eviction.",[22,3749,3750],{},"Practical limit: 80-90% VRAM usable; >80% crashes from host RAM exhaustion in CUDA Graph compilation (needs headroom for metadata\u002Fworkspace). Rule: Budget 80% for weights\u002FKV, reserve 20%.",[22,3752,3753],{},"\"By eliminating the contiguity constraint, PagedAttention pushed memory utilization from the 20 to 40% range up to over 96% in optimized deployments.\"",[17,3755,3757],{"id":3756},"continuous-batching-and-low-overhead-scheduling-maximize-gpu-utilization","Continuous Batching and Low-Overhead Scheduling Maximize GPU Utilization",[22,3759,3760],{},"Static batching processes full batches to slowest request's end, padding short ones and blocking queue—tail latency spikes. Continuous batching (vLLM\u002FSGLang) swaps per-token: evict completes, admit waits if KV slots free, no idle GPU.",[22,3762,3763],{},"Scheduler overhead grows with speed\u002Fbatch size; Python (vLLM) flexible but slower. LMDeploy's C++ TurboMind hits microsecond precision, 29% higher throughput than vLLM on H100s via compiled batch mgmt\u002Fmemory\u002Frequests. vLLM wins on ecosystem\u002Fflexibility for most; TurboMind for peak high-concurrency.",[22,3765,3766],{},"\"The GPU never processes a completed request for even one unnecessary iteration, and new requests begin generation as soon as a slot opens.\"",[17,3768,3770],{"id":3769},"speculative-decoding-accelerates-autoregression-2-65x","Speculative Decoding Accelerates Autoregression 2-6.5x",[22,3772,3773],{},"Each decode step moves 10s GB for 70B models. Speculative decoding drafts N tokens cheap\u002Ffast, verifies parallel in target model (preserves distribution exactly).",[22,3775,3776],{},"Traditional: Separate small draft model (e.g., 1B for 70B), 40-60% acceptance, 2-3x speedup, but VRAM\u002Fsync overhead and weaker drafts.",[22,3778,3779],{},"EAGLE-3 integrates autoregressive heads on target's hidden states (multi-layer fusion), seeing rich embeddings for superior drafts. Dynamic tree verification (not linear seq). 3-6.5x speedup (5.6x on Vicuna-13B vs. vanilla, 1.8x vs. EAGLE-1), 20-40% over EAGLE-2. Task-variant (high on code\u002Ftemplates, lower math). \"The most effective inference optimizations are not the ones that work around the model. They are the ones that work with the model’s own internal structure.\"",[17,3781,3783],{"id":3782},"multi-lora-servings-cache-interference-demands-unified-management","Multi-LoRA Serving's Cache Interference Demands Unified Management",[22,3785,3786],{},"Single base in VRAM, swap tiny LoRA adapters (hundreds MB) for variants (support\u002Fcode\u002Fsummarization). But KV cache adapter-specific: eviction orphans invalid cache (up to 46.5% in vLLM), bloating TTFT.",[22,3788,3789],{},"FastLibra (ELORA) links adapters\u002FKV in shared tree pool; evict pairs via TTFT-impact cost model (retain hot adapters). 63.4% TTFT reduction, 1.7x peak throughput vs. vLLM.",[22,3791,3792],{},"\"A KV cache entry is only valid for the specific adapter that produced it... Experimental data shows that vLLM can reach an invalid KV cache rate of up to 46.5% in high-churn multi-LoRA workloads.\"",[17,3794,3796],{"id":3795},"prefill-decode-disaggregation-quantization-and-engine-landscape","Prefill-Decode Disaggregation, Quantization, and Engine Landscape",[22,3798,3799],{},"Prefill (parallel, compute) and decode (serial, memory) disaggregate to specialized hardware: H100s for decode bandwidth, A100s for prefill FLOPS. Quantization (INT4\u002FINT8) shrinks weights 4-8x with \u003C1% perf loss, but structured outputs need careful handling (e.g., logit biasing). Engines: vLLM (PagedAttention, Python-flex), SGLang (RadixAttention), LMDeploy (TurboMind C++ speed), hardware reality favors HBM-heavy GPUs like H100\u002FH200.",[17,3801,3803],{"id":3802},"key-takeaways","Key Takeaways",[3805,3806,3807,3811,3814,3817,3820,3823,3826,3829,3832],"ul",{},[3808,3809,3810],"li",{},"Target 80% VRAM utilization max; reserve 20% for CUDA Graphs\u002Fhost overhead.",[3808,3812,3813],{},"Deploy PagedAttention (vLLM) for 2-4x baseline throughput via dynamic paging.",[3808,3815,3816],{},"Use continuous batching to eliminate static padding\u002Ftail latency.",[3808,3818,3819],{},"Integrate EAGLE-3-style heads for 3-6.5x speculative gains over separate drafts.",[3808,3821,3822],{},"For multi-LoRA, adopt FastLibra to evict adapter-KV pairs, cutting TTFT 63%.",[3808,3824,3825],{},"Prioritize HBM bandwidth over FLOPS; disaggregate prefill\u002Fdecode if scaling.",[3808,3827,3828],{},"Benchmark engines: vLLM for broad use, LMDeploy\u002FSGLang for peak H100 perf.",[3808,3830,3831],{},"Quantize aggressively (INT4) post-training, validate structured outputs.",[3808,3833,3834],{},"Reuse prefixes with RadixAttention for 6.4x in chats\u002Fagents.",{"title":40,"searchDepth":41,"depth":41,"links":3836},[3837,3838,3839,3840,3841,3842,3843],{"id":3730,"depth":41,"text":3731},{"id":3740,"depth":41,"text":3741},{"id":3756,"depth":41,"text":3757},{"id":3769,"depth":41,"text":3770},{"id":3782,"depth":41,"text":3783},{"id":3795,"depth":41,"text":3796},{"id":3802,"depth":41,"text":3803},[],{"content_references":3846,"triage":3847},[],{"relevance":66,"novelty":67,"quality":67,"actionability":67,"composite":68,"reasoning":3848},"Category: AI & LLMs. The article provides in-depth insights into optimizing LLM inference, addressing specific challenges like memory-bound latency and KV cache utilization, which are critical for developers building AI-powered products. It offers actionable techniques such as PagedAttention and continuous batching that can be directly applied in production environments.","\u002Fsummaries\u002F7b130eb6998f566d-scaling-llm-inference-kv-cache-batching-spec-decod-summary","2026-04-15 20:01:01","2026-04-16 03:18:51",{"title":3720,"description":40},{"loc":3849},"7b130eb6998f566d","Towards AI","https:\u002F\u002Fpub.towardsai.net\u002Fllm-inference-infrastructure-from-scratch-how-to-fine-tune-correctly-part-7-5df0f1494b97?source=rss----98111c9905da---4","summaries\u002F7b130eb6998f566d-scaling-llm-inference-kv-cache-batching-spec-decod-summary",[3859,81,82,3860],"llm","software-engineering","Production LLM serving shifts from training's throughput focus to inference's memory-bound latency challenges, solved by PagedAttention (96% util), continuous batching, EAGLE-3 (up to 6.5x speedup), and FastLibra for multi-LoRA (63% TTFT cut).",[81,82,3860],"7xjH9_abrODqr5NEFl2e2fJrbyQNbF2IibSYkdQJZGc",{"id":3865,"title":3866,"ai":3867,"body":3872,"categories":3981,"created_at":48,"date_modified":48,"description":40,"extension":49,"faq":48,"featured":50,"kicker_label":48,"meta":3982,"navigation":70,"path":3983,"published_at":3984,"question":48,"scraped_at":48,"seo":3985,"sitemap":3986,"source_id":3987,"source_name":3988,"source_type":77,"source_url":3989,"stem":3990,"tags":3991,"thumbnail_url":48,"tldr":3994,"tweet":48,"unknown_tags":3995,"__hash__":3996},"summaries\u002Fsummaries\u002Fai-chokepoints-chips-power-reshape-global-race-summary.md","AI Chokepoints: Chips, Power Reshape Global Race",{"provider":7,"model":8,"input_tokens":3868,"output_tokens":3869,"processing_time_ms":3870,"cost_usd":3871},8929,2984,19191,0.00325515,{"type":14,"value":3873,"toc":3975},[3874,3878,3881,3884,3887,3890,3894,3897,3900,3903,3906,3909,3912,3916,3919,3922,3925,3928,3931,3934,3938,3941,3944,3947,3949],[17,3875,3877],{"id":3876},"_2026-supply-chain-crises-hit-ai-hardware-hard","2026 Supply Chain Crises Hit AI Hardware Hard",[22,3879,3880],{},"AI production faces immediate \"RAMageddon\" from structural DRAM and HBM shortages, exacerbated by a helium crisis tied to the Iran War. Helium, essential for over 20 semiconductor fab steps, sees Qatar (34% global supply) blocked via Strait of Hormuz closure. Ras Laffan facility alone provides 30-33% of world helium; South Korea, sourcing 64.7% from Qatar and producing 60%+ of global memory via Samsung\u002FSK Hynix, is hit hardest. This amplifies HBM bottlenecks—3D-stacked DRAM skyscrapers for Nvidia AI GPUs—driving prices up as silicon wafer capacity reallocates to AI over consumer uses. TSMC remains the core GPU bottleneck, but helium scarcity slows everything upstream.",[22,3882,3883],{},"Datacenter buildouts halved in Q4 2025 per Wood Mackenzie data: of 241GW disclosed capacity, only 33% is under active development. Factors include community opposition, speculative large projects, and grid limits. Paul Kedrosky charts show sharp slowdowns; Ed Zitron calls out AI industry hype amid reality bites.",[22,3885,3886],{},"In chips news, ARM breaks 35-year IP-only model with AGI CPU, selling physical chips to Meta, OpenAI, SAP, Cloudflare. Designed for agentic AI orchestration—autonomous reasoning\u002Facting systems—announced March 24, 2026, in ARM Everywhere keynote.",[22,3888,3889],{},"\"The Iran War has created a choke point in the supply of helium... used in more than 20 steps of semiconductor fabrication.\" — Nathan Warren, Exponential View.",[17,3891,3893],{"id":3892},"physical-and-institutional-constraints-override-software-diffusion","Physical and Institutional Constraints Override Software Diffusion",[22,3895,3896],{},"Past decade's AI relied on fast-diffusing inputs: algorithms, papers, open-source, talent. Microsoft AI Diffusion Report shows AI spreading slower than internet\u002Fmobile but faster than many techs—until now. Frontier AI hinges on data centers converting electricity to compute at scale, bound by unevenly distributed chips, power, capital, institutions.",[22,3898,3899],{},"Harvard Belfer Center's National AI Capability Index decomposes by compute, data, algorithms, human capital, resources, regulation, performance—revealing US dominance, uneven global spread. Chokepoints make frontier capability geographically concentrated where silicon\u002Fpower\u002Ffinance\u002Fpolitics align.",[22,3901,3902],{},"\"For much of the past decade, AI progress appeared to be driven by ideas that diffused easily across borders. That model no longer holds. Today, frontier artificial intelligence is constrained by geopolitical chokepoints: access to advanced chips, the ability to deliver large amounts of electricity quickly, and the capital and institutions required to build and operate massive data centers.\"",[22,3904,3905],{},"Software efficiency improves (OpenAI: 10x compute reduction 2012-2022; Epoch AI: LLM compute halves every 8 months, beating Moore's Law), but diffuses globally via papers\u002Fframeworks\u002Ftalent. Thus, it's no chokepoint—everyone advances together. Instead, it spurs Jevons paradox: efficiency lowers intelligence cost, fueling more compute spend (Bain\u002FIMF: \"resource race\" outpaces gains).",[22,3907,3908],{},"Scaling laws persist: Epoch Capabilities Index shows added compute yields frontier gains. DeepSeek (China) stress-tests: algorithmic wins spur spending, not less (Epoch: progress likely increases compute demand).",[22,3910,3911],{},"Training compute for top models grows exponentially per Epoch AI trends.",[17,3913,3915],{"id":3914},"power-emerges-as-ultimate-scaling-limiter","Power Emerges as Ultimate Scaling Limiter",[22,3917,3918],{},"With chips secured, electricity dictates growth. Data centers hit 415 TWh in 2024 (1.5% global); IEA projects 700-1,700 TWh by 2035, doubling\u002Ftripling via sustained AI loads. Frontier clusters draw 100-500MW continuously—like heavy industry or mid-size cities (300MW site: 2.6 TWh\u002Fyear).",[22,3920,3921],{},"2025-2026 projects scale to gigawatts: xAI Colossus, OpenAI Stargate, Meta Hyperion campuses span urban-scale land, per Epoch AI maps. Goldman Sachs: AI drives 165% data center power jump by 2030.",[22,3923,3924],{},"Cheaper solar\u002Fbatteries help, but timing kills: need MWs now, on AI cycles—not utility years. Constraints: permitting, grid queues, transmission. Modular solar\u002Fstorage wins for speed (faster than thermal\u002Fgrid), prioritizing velocity over price. \"Delays matter more than electricity prices. A year without power means a year without training runs.\"",[22,3926,3927],{},"US edges via faster permitting\u002Fmodular tech; AI accelerates energy transition ironically.",[22,3929,3930],{},"![Projected power growth of leading AI data centers... approaching the electricity demand of major cities. — Epoch AI](image placeholder)",[22,3932,3933],{},"\"Electricity becomes the principal variable that determines how large AI systems can grow.\"",[17,3935,3937],{"id":3936},"implications-concentrated-ai-power-reshapes-geopolitics","Implications: Concentrated AI Power Reshapes Geopolitics",[22,3939,3940],{},"Chips\u002Fpower\u002Fcapital concentrate frontier AI in US (Belfer index lead), despite China algorithmic pushes like DeepSeek. Export controls stockpile-able; power not. Software accelerates all, but physicals decide leaders.",[22,3942,3943],{},"\"Software efficiency continues to improve, but it accelerates competition rather than leveling it. As a result, frontier AI capability is becoming geographically concentrated.\"",[22,3945,3946],{},"Epoch AI power projections: from near-zero to GW-scale by 2026-27. Bain: meet insatiable compute via scale. IMF: AI-led resource race.",[22,3948,3803],{},[3805,3950,3951,3954,3957,3960,3963,3966,3969,3972],{},[3808,3952,3953],{},"Monitor helium\u002FHBM supply: Qatar disruptions (30%+ global) hit fabs hardest; diversify or stockpile for AI GPU builds.",[3808,3955,3956],{},"Factor power timelines into infra plans: Prioritize sites with fast permitting\u002Fmodular solar+batteries over cheap long-term energy.",[3808,3958,3959],{},"Expect Jevons-driven compute explosion: Efficiency gains mean more spending—budget for 2-3x power by 2030.",[3808,3961,3962],{},"Bet on concentrated leaders: US wins short-term via institutions\u002Fpower; track xAI\u002FOpenAI\u002FMeta campuses.",[3808,3964,3965],{},"Agentic AI hardware shift: ARM's AGI CPU signals orchestration chips for autonomous agents—prototype with early access.",[3808,3967,3968],{},"Avoid hype slowdowns: Datacenter pipelines halved; validate 33% active dev before scaling commitments.",[3808,3970,3971],{},"Stress-test like DeepSeek: Use efficiency to spend more compute, not save—push frontiers despite costs.",[3808,3973,3974],{},"Geopolitics matters: Iran War\u002FStrait Hormuz shows non-China risks; model supply chains end-to-end.",{"title":40,"searchDepth":41,"depth":41,"links":3976},[3977,3978,3979,3980],{"id":3876,"depth":41,"text":3877},{"id":3892,"depth":41,"text":3893},{"id":3914,"depth":41,"text":3915},{"id":3936,"depth":41,"text":3937},[115],{},"\u002Fsummaries\u002Fai-chokepoints-chips-power-reshape-global-race-summary","2026-04-08 21:21:17",{"title":3866,"description":40},{"loc":3983},"e1894582f1e9e5eb","AI Supremacy","https:\u002F\u002Funknown","summaries\u002Fai-chokepoints-chips-power-reshape-global-race-summary",[3992,3993,81,82],"machine-learning","agents","Frontier AI shifts from diffusible software to physical chokepoints in chips, helium, HBM\u002FDRAM, power delivery, concentrating capability in few geographies like the US.",[81,82],"5eH_135QVpX2ORp8Q8GkESa7NwF3HEpCQHvFr_UalEk",{"id":3998,"title":3999,"ai":4000,"body":4005,"categories":4070,"created_at":48,"date_modified":48,"description":40,"extension":49,"faq":48,"featured":50,"kicker_label":48,"meta":4071,"navigation":70,"path":4107,"published_at":4108,"question":48,"scraped_at":4109,"seo":4110,"sitemap":4111,"source_id":4112,"source_name":4113,"source_type":77,"source_url":4114,"stem":4115,"tags":4116,"thumbnail_url":48,"tldr":4118,"tweet":48,"unknown_tags":4119,"__hash__":4120},"summaries\u002Fsummaries\u002F6b8835c7aeff291e-ai-scales-disordered-human-values-not-truth-summary.md","AI Scales Disordered Human Values, Not Truth",{"provider":7,"model":8,"input_tokens":4001,"output_tokens":4002,"processing_time_ms":4003,"cost_usd":4004},5849,2343,28581,0.00231865,{"type":14,"value":4006,"toc":4065},[4007,4011,4039,4042,4046,4049,4052,4056,4059,4062],[17,4008,4010],{"id":4009},"augustines-diagnosis-systems-fail-from-misdirected-loves","Augustine's Diagnosis: Systems Fail from Misdirected Loves",[22,4012,4013,4014,4018,4019,4022,4023,4026,4027,4030,4031,4034,4035,4038],{},"Human systems collapse not from poor execution but misordered desires—what Augustine calls ",[4015,4016,4017],"em",{},"ordo amoris",". In ",[4015,4020,4021],{},"Confessions",", he shows desire shapes what we build; societies reflect collective loves. The City of Man prioritizes self-love and ",[4015,4024,4025],{},"libido dominandi"," (mastery drive), chasing unstable goods like security, leading to inherent instability. The City of God orients toward divine order via rightly ordered love. Tools follow ",[4015,4028,4029],{},"uti"," (use as means, from ",[4015,4032,4033],{},"On Christian Doctrine",") vs. ",[4015,4036,4037],{},"frui"," (enjoy as end)—AI errs by blurring this, treating utility as authority. Even secularly, this mirrors bounded rationality: no system self-justifies its value hierarchy, whether from cognitive limits or original sin.",[22,4040,4041],{},"This creates a structural gap: optimization needs a prior 'good' definition, always partial and contested. Builders ignore this at peril—AI doesn't access fundamental truth, only scales encoded values.",[17,4043,4045],{"id":4044},"ais-false-promise-efficiency-masks-distortion","AI's False Promise: Efficiency Masks Distortion",[22,4047,4048],{},"AI intensifies the problem by optimizing flawed inputs at scale. Hiring algorithms narrow 'qualified' to keywords; recommendation systems redefine relevance; risk models formalize biases. MIT Technology Review coverage shows AI doesn't eliminate bias but embeds it objectively. Generative tools or AGI pursuits assume more intelligence resolves value disputes—it doesn't, just amplifies priors. Engagement as 'good' maximizes attention; efficiency sacrifices depth. Outputs become conclusions, narrowing human perception: the tool frames reality, not windows it.",[22,4050,4051],{},"Innovation feels precise via metrics, but unexamined norms persist. Consistent results signal stability, not legitimacy—a well-oiled City of Man machine.",[17,4053,4055],{"id":4054},"remedies-for-builders-judgment-over-automation","Remedies for Builders: Judgment Over Automation",[22,4057,4058],{},"Restore deliberation: AI outputs are inputs to human reasoning, not finals. Structure orgs so judgment stays authoritative—e.g., review hiring scores manually.",[22,4060,4061],{},"Expose values: Surface embedded priorities as political choices open to contestation, per algorithmic accountability research. Name assumptions in models (e.g., what 'qualified' means) for revision.",[22,4063,4064],{},"Cultivate institutional humility: Audit if outputs align with right goals, not just stated ones. Efficiency doesn't validate ends. Result: AI aids without substituting moral seriousness, preserving systems from disordered orientations.",{"title":40,"searchDepth":41,"depth":41,"links":4066},[4067,4068,4069],{"id":4009,"depth":41,"text":4010},{"id":4044,"depth":41,"text":4045},{"id":4054,"depth":41,"text":4055},[47],{"content_references":4072,"triage":4103},[4073,4077,4080,4082,4086,4089,4093,4096,4099],{"type":4074,"title":4021,"author":4075,"url":4076,"context":57},"book","Saint Augustine","https:\u002F\u002Fwww.newadvent.org\u002Ffathers\u002F1101.htm",{"type":4074,"title":4078,"author":4075,"url":4079,"context":57},"City of God","https:\u002F\u002Fwww.newadvent.org\u002Ffathers\u002F1201.htm",{"type":4074,"title":4033,"author":4075,"url":4081,"context":57},"https:\u002F\u002Fwww.newadvent.org\u002Ffathers\u002F1202.htm",{"type":59,"title":4083,"url":4084,"context":4085},"Saint Augustine of Hippo","https:\u002F\u002Fplato.stanford.edu\u002Fentries\u002Faugustine\u002F","mentioned",{"type":59,"title":4087,"url":4088,"context":4085},"Artificial general intelligence","https:\u002F\u002Fwww.ibm.com\u002Fthink\u002Ftopics\u002Fartificial-general-intelligence",{"type":54,"title":4090,"publisher":4091,"url":4092,"context":57},"AI bias explained","MIT Technology Review","https:\u002F\u002Fwww.technologyreview.com\u002F2020\u002F02\u002F14\u002F844765\u002Fai-bias-explained\u002F",{"type":59,"title":4094,"url":4095,"context":57},"Bounded Rationality","https:\u002F\u002Fplato.stanford.edu\u002Fentries\u002Fbounded-rationality\u002F",{"type":59,"title":4097,"url":4098,"context":4085},"What did St. Augustine say about original sin?","https:\u002F\u002Fuscatholic.org\u002Farticles\u002F202411\u002Fwhat-did-st-augustine-say-about-original-sin\u002F",{"type":54,"title":4100,"publisher":4101,"url":4102,"context":57},"Algorithmic Accountability","Data & Society","https:\u002F\u002Fdatasociety.net\u002Flibrary\u002Falgorithmic-accountability\u002F",{"relevance":4104,"novelty":4104,"quality":67,"actionability":4104,"composite":4105,"reasoning":4106},3,3.25,"Category: product-strategy. The article discusses the implications of AI on human values and decision-making, which is relevant to product strategy in AI development. It provides some insights into the risks of automation but lacks concrete, actionable steps for builders to implement in their workflows.","\u002Fsummaries\u002F6b8835c7aeff291e-ai-scales-disordered-human-values-not-truth-summary","2026-05-06 10:30:00","2026-05-08 15:34:04",{"title":3999,"description":40},{"loc":4107},"6b8835c7aeff291e","UX Collective","https:\u002F\u002Fuxdesign.cc\u002Fst-augustine-and-ais-false-promise-4f67c75b3275?source=rss----138adf9c44c---4","summaries\u002F6b8835c7aeff291e-ai-scales-disordered-human-values-not-truth-summary",[4117,81],"product-strategy","AI optimizes for predefined 'good' but embeds unstable human values, amplifying biases; builders must prioritize human judgment over automation to avoid mistaking tools for ends.",[81],"h3LCKc4HyC47Ey6ITApf6mImvxikRwSDsCyTF2tC9og"]