{"id":12519,"date":"2024-05-22T18:16:07","date_gmt":"2024-05-22T18:16:07","guid":{"rendered":"https:\/\/dailyai.com\/?p=12519"},"modified":"2024-05-24T10:16:44","modified_gmt":"2024-05-24T10:16:44","slug":"decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box","status":"publish","type":"post","link":"https:\/\/dailyai.com\/it\/2024\/05\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\/","title":{"rendered":"Decodificare la mente dell'intelligenza artificiale: Ricercatori antropici scrutano all'interno della \"scatola nera\""},"content":{"rendered":"<p><b><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\">Anthropic<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span> I ricercatori hanno identificato con successo milioni di concetti all'interno di <span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\">Claude<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span> Sonnet, uno dei loro LLM avanzati.<\/b><\/p>\n<p><span style=\"font-weight: 400;\">I modelli di intelligenza artificiale sono spesso considerati scatole nere, ovvero non \u00e8 possibile \"vedere\" al loro interno per capire esattamente come funzionano.\u00a0<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Quando si fornisce un input a un LLM, questo genera una risposta, ma il ragionamento dietro le sue scelte non \u00e8 chiaro. <\/span><\/p>\n<p><span style=\"font-weight: 400;\">L'input entra e l'output esce, e nemmeno gli stessi sviluppatori di IA capiscono veramente cosa succede all'interno di quella \"scatola\".\u00a0<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Le reti neurali creano le proprie rappresentazioni interne delle informazioni quando mappano gli input agli output durante l'addestramento dei dati.\u00a0<\/span><span style=\"font-weight: 400;\">Gli elementi costitutivi di questo processo, chiamati \"attivazioni neuronali\", sono rappresentati da valori numerici. <\/span><\/p>\n<p><span style=\"font-weight: 400;\">Ogni concetto \u00e8 distribuito su pi\u00f9 neuroni e ogni neurone contribuisce a rappresentare pi\u00f9 concetti, rendendo difficile la mappatura dei concetti direttamente ai singoli neuroni.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Si tratta di una situazione ampiamente analoga a quella del nostro cervello umano. Cos\u00ec come il nostro cervello elabora gli input sensoriali e genera pensieri, comportamenti e ricordi, i miliardi, o addirittura trilioni, di processi che stanno alla base di queste funzioni rimangono principalmente sconosciuti alla scienza.<\/span><\/p>\n<p><a href=\"https:\/\/www.anthropic.com\/research\/mapping-mind-language-model\"><span style=\"font-weight: 400;\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\">Anthropic<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span>studio<\/span><\/a><span style=\"font-weight: 400;\"> tenta di vedere all'interno della scatola nera dell'intelligenza artificiale con una tecnica chiamata \"apprendimento a dizionario\".\u00a0<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Ci\u00f2 comporta la scomposizione di modelli complessi in un modello di intelligenza artificiale in blocchi lineari o \"atomi\" che abbiano un senso intuitivo per gli esseri umani.<\/span><\/p>\n<h2>Mappare gli LLM con l'apprendimento dei dizionari<\/h2>\n<p><span style=\"font-weight: 400;\">Nell'ottobre 2023, <span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\">Anthropic<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span> ha applicato questo metodo a un piccolo modello linguistico \"giocattolo\" e ha trovato caratteristiche coerenti corrispondenti a concetti come il testo maiuscolo, le sequenze di DNA, i cognomi nelle citazioni, i sostantivi matematici o gli argomenti delle funzioni nel codice Python.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Quest'ultimo studio ha ampliato la tecnica in modo da poterla utilizzare, in questo caso, per i modelli linguistici di IA pi\u00f9 grandi, <span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\">Anthropic<\/span><\/span><\/span><\/span>'s <span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\">Claude<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span> 3 Sonetto.\u00a0<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Ecco una descrizione passo passo di come ha funzionato lo studio:<\/span><\/p>\n<h4><b>Identificare i modelli con l'apprendimento dei dizionari<\/b><\/h4>\n<p><span style=\"font-weight: 400;\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\">Anthropic<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span> ha utilizzato l'apprendimento a dizionario per analizzare le attivazioni dei neuroni in vari contesti e identificare modelli comuni. <\/span><\/p>\n<p><span style=\"font-weight: 400;\">L'apprendimento dizionario raggruppa queste attivazioni in un insieme pi\u00f9 piccolo di \"caratteristiche\" significative, che rappresentano concetti di livello superiore appresi dal modello. <\/span><\/p>\n<p><span style=\"font-weight: 400;\">Identificando queste caratteristiche, i ricercatori possono capire meglio come il modello elabora e rappresenta le informazioni.<\/span><\/p>\n<h4><b>Estrazione delle caratteristiche dallo strato intermedio<\/b><\/h4>\n<p><span style=\"font-weight: 400;\">I ricercatori si sono concentrati sullo strato intermedio della <span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\">Claude<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span> 3.0 Sonnet, che funge da punto critico nella pipeline di elaborazione del modello. <\/span><\/p>\n<p><span style=\"font-weight: 400;\">Applicando l'apprendimento a dizionario a questo livello si estraggono milioni di caratteristiche che catturano le rappresentazioni interne del modello e i concetti appresi in questa fase. <\/span><\/p>\n<p><span style=\"font-weight: 400;\">L'estrazione di caratteristiche dallo strato intermedio consente ai ricercatori di esaminare la comprensione delle informazioni da parte del modello. <em>dopo<\/em> ha elaborato l'input <em>prima<\/em> generando l'output finale.<\/span><\/p>\n<h4><b>Scoprire concetti diversi e astratti<\/b><\/h4>\n<p><span style=\"font-weight: 400;\">Le caratteristiche estratte hanno rivelato un'ampia gamma di concetti appresi da <span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\">Claude<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span>da entit\u00e0 concrete come citt\u00e0 e persone a nozioni astratte relative a campi scientifici e sintassi di programmazione. <\/span><\/p>\n<p><span style=\"font-weight: 400;\">\u00c8 interessante notare che le caratteristiche sono risultate multimodali, rispondendo a input sia testuali che visivi, indicando che il modello pu\u00f2 apprendere e rappresentare concetti attraverso diverse modalit\u00e0. <\/span><\/p>\n<p><span style=\"font-weight: 400;\">Inoltre, le caratteristiche multilingue suggeriscono che il modello \u00e8 in grado di cogliere i concetti espressi in diverse lingue.<\/span><\/p>\n<p><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"598\" class=\"wp-image-12521 size-large\" src=\"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/c896a301ad3d1ef4237cb05b68d78b467c444097-2200x1284-1-1024x598.webp\" alt=\"&lt;span class=\" srcset=\"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/c896a301ad3d1ef4237cb05b68d78b467c444097-2200x1284-1-1024x598.webp 1024w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/c896a301ad3d1ef4237cb05b68d78b467c444097-2200x1284-1-300x175.webp 300w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/c896a301ad3d1ef4237cb05b68d78b467c444097-2200x1284-1-768x448.webp 768w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/c896a301ad3d1ef4237cb05b68d78b467c444097-2200x1284-1-1536x896.webp 1536w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/c896a301ad3d1ef4237cb05b68d78b467c444097-2200x1284-1-2048x1195.webp 2048w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/c896a301ad3d1ef4237cb05b68d78b467c444097-2200x1284-1-18x12.webp 18w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/c896a301ad3d1ef4237cb05b68d78b467c444097-2200x1284-1-60x35.webp 60w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/p>\n<h3><b>Analizzare l'organizzazione dei concetti<\/b><\/h3>\n<p><span style=\"font-weight: 400;\">Per capire come il modello organizza e mette in relazione i diversi concetti, i ricercatori hanno analizzato la somiglianza tra le caratteristiche in base ai loro modelli di attivazione. <\/span><\/p>\n<p><span style=\"font-weight: 400;\">Hanno scoperto che le caratteristiche che rappresentano concetti correlati tendevano a raggrupparsi insieme. Ad esempio, le caratteristiche associate alle citt\u00e0 o alle discipline scientifiche mostravano una maggiore somiglianza tra loro rispetto alle caratteristiche che rappresentavano concetti non correlati. <\/span><\/p>\n<p><span style=\"font-weight: 400;\">Ci\u00f2 suggerisce che l'organizzazione interna dei concetti del modello si allinea, in una certa misura, alle intuizioni umane sulle relazioni concettuali.<\/span><\/p>\n<figure id=\"attachment_12520\" aria-describedby=\"caption-attachment-12520\" style=\"width: 1024px\" class=\"wp-caption aligncenter\"><img loading=\"lazy\" decoding=\"async\" class=\"wp-image-12520 size-large\" src=\"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/fe4d42c004bf43efda0f5921adfedd2f8f42e417-2200x2140-1-1024x996.webp\" alt=\"antropico\" width=\"1024\" height=\"996\" srcset=\"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/fe4d42c004bf43efda0f5921adfedd2f8f42e417-2200x2140-1-1024x996.webp 1024w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/fe4d42c004bf43efda0f5921adfedd2f8f42e417-2200x2140-1-300x292.webp 300w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/fe4d42c004bf43efda0f5921adfedd2f8f42e417-2200x2140-1-768x747.webp 768w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/fe4d42c004bf43efda0f5921adfedd2f8f42e417-2200x2140-1-1536x1494.webp 1536w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/fe4d42c004bf43efda0f5921adfedd2f8f42e417-2200x2140-1-2048x1992.webp 2048w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/fe4d42c004bf43efda0f5921adfedd2f8f42e417-2200x2140-1-12x12.webp 12w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/fe4d42c004bf43efda0f5921adfedd2f8f42e417-2200x2140-1-60x58.webp 60w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/fe4d42c004bf43efda0f5921adfedd2f8f42e417-2200x2140-1-24x24.webp 24w, https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/fe4d42c004bf43efda0f5921adfedd2f8f42e417-2200x2140-1-48x48.webp 48w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><figcaption id=\"caption-attachment-12520\" class=\"wp-caption-text\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\">Anthropic<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span> \u00e8 riuscito a mappare concetti astratti come \"conflitto interiore\". Fonte: <a href=\"https:\/\/www.anthropic.com\/research\/mapping-mind-language-model\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\">Anthropic<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/a>.<\/figcaption><\/figure>\n<h4><b>Verifica delle caratteristiche<\/b><\/h4>\n<p><span style=\"font-weight: 400;\">Per confermare che le caratteristiche identificate influenzano direttamente il comportamento e gli output del modello, i ricercatori hanno condotto esperimenti di \"feature steering\". <\/span><\/p>\n<p><span style=\"font-weight: 400;\">Si trattava di amplificare o sopprimere selettivamente l'attivazione di caratteristiche specifiche durante l'elaborazione del modello e di osservare l'impatto sulle sue risposte. <\/span><\/p>\n<p><span style=\"font-weight: 400;\">Manipolando le singole caratteristiche, i ricercatori hanno potuto stabilire un legame diretto tra le singole caratteristiche e il comportamento del modello. Ad esempio, l'amplificazione di una caratteristica relativa a una citt\u00e0 specifica ha fatto s\u00ec che il modello generasse output basati sulla citt\u00e0, anche in contesti irrilevanti.<\/span><\/p>\n<p>Leggi lo studio completo <a href=\"https:\/\/transformer-circuits.pub\/2024\/scaling-monosemanticity\/index.html\" target=\"_blank\" rel=\"noopener\">qui<\/a>.<\/p>\n<p><iframe loading=\"lazy\" title=\"Apprendimento del dizionario su Claude 3 Sonnet\" width=\"1080\" height=\"608\" src=\"https:\/\/www.youtube.com\/embed\/CJIbCV92d88?feature=oembed\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share\" referrerpolicy=\"strict-origin-when-cross-origin\" allowfullscreen><\/iframe><\/p>\n<h2>Perch\u00e9 l'interpretabilit\u00e0 \u00e8 fondamentale per la sicurezza dell'IA<\/h2>\n<p><span style=\"font-weight: 400;\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\">Anthropic<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span>La ricerca dell'autore \u00e8 fondamentalmente rilevante per l'interpretabilit\u00e0 dell'IA e, per estensione, per la sicurezza.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Capire come i LLM elaborano e rappresentano le informazioni aiuta i ricercatori a capire e a mitigare i rischi. Il <\/span><span style=\"font-weight: 400;\">pone le basi per lo sviluppo di sistemi di IA pi\u00f9 trasparenti e spiegabili.\u00a0<\/span><\/p>\n<p>Come <span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\">Anthropic<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span> Speriamo che noi e altri possano utilizzare queste scoperte per rendere i modelli pi\u00f9 sicuri. Ad esempio, potrebbe essere possibile utilizzare le tecniche qui descritte per monitorare i sistemi di IA per alcuni comportamenti pericolosi (come ingannare l'utente), per indirizzarli verso risultati desiderabili (debiasing) o per eliminare completamente alcuni argomenti pericolosi\".<\/p>\n<p><span style=\"font-weight: 400;\">La comprensione del comportamento dell'IA diventa fondamentale in quanto l'IA diventa onnipresente nei processi decisionali critici in settori quali l'assistenza sanitaria, la finanza e la giustizia penale. Inoltre, aiuta a scoprire la causa principale di <a href=\"https:\/\/dailyai.com\/it\/2023\/06\/anthropic-releases-paper-highlighting-the-potential-bias-of-large-language-models\/\">sbieco<\/a>allucinazioni e altri comportamenti indesiderati o imprevedibili.\u00a0<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Ad esempio, un <\/span><a href=\"https:\/\/dailyai.com\/it\/2024\/01\/researchers-aim-to-determine-how-black-box-drug-discovery-models-work\/\"><span style=\"font-weight: 400;\">studio recente<\/span><\/a><span style=\"font-weight: 400;\"> dell'Universit\u00e0 di Bonn ha scoperto come le reti neurali a grafo (GNN) utilizzate per la scoperta di farmaci si basino in larga misura sul richiamo di somiglianze dai dati di addestramento piuttosto che sull'apprendimento reale di nuove interazioni chimiche complesse. <\/span><\/p>\n<p><span style=\"font-weight: 400;\">Ci\u00f2 rende difficile capire come questi modelli determinino esattamente i nuovi composti di interesse.<\/span><\/p>\n<p><span style=\"font-weight: 400;\">L'anno scorso, il <\/span><a href=\"https:\/\/dailyai.com\/it\/2023\/09\/the-uk-government-wants-to-see-inside-ais-black-box\/\"><span style=\"font-weight: 400;\">Il governo britannico ha negoziato<\/span><\/a><span style=\"font-weight: 400;\"> con grandi colossi tecnologici come <span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\">OpenAI<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span> e <span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\">DeepMind<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span>che cercano di accedere ai processi decisionali interni dei loro sistemi di intelligenza artificiale.\u00a0<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Regolamenti come il <a href=\"https:\/\/dailyai.com\/it\/2024\/03\/the-eu-ai-act-passed-in-a-landslide-and-will-come-into-force-this-year\/\">Legge sull'intelligenza artificiale dell'UE<\/a> far\u00e0 pressione sulle aziende di IA affinch\u00e9 siano pi\u00f9 trasparenti, anche se i segreti commerciali sembrano destinati a rimanere sotto chiave.\u00a0<\/span><\/p>\n<p><span style=\"font-weight: 400;\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\">Anthropic<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span>La ricerca dell'Istituto offre un'idea di ci\u00f2 che si trova all'interno della scatola, \"mappando\" le informazioni all'interno del modello.\u00a0<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Tuttavia, la verit\u00e0 \u00e8 che questi modelli sono cos\u00ec vasti che, per <span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\"><span class=\"noTranslate\" data-no-translation=\"\">Anthropic<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span>Per stessa ammissione dell'autore: \"Pensiamo che sia abbastanza probabile che ci manchino ordini di grandezza e che se volessimo ottenere tutte le caratteristiche - in tutti i livelli! - dovremmo utilizzare una quantit\u00e0 di calcolo molto superiore a quella necessaria per addestrare i modelli sottostanti\".<\/span><\/p>\n<p><span style=\"font-weight: 400;\">\u00c8 un punto interessante: il reverse engineering di un modello \u00e8 pi\u00f9 complesso dal punto di vista computazionale rispetto alla progettazione del modello stesso.\u00a0<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Ricorda i progetti di neuroscienza estremamente costosi come il <a href=\"https:\/\/dailyai.com\/it\/2023\/08\/exploring-impact-of-the-human-brain-project-hbp-on-science-and-ai\/\">Progetto Cervello Umano (HBP)<\/a>che ha investito miliardi nella mappatura del nostro cervello umano, ma alla fine ha fallito.\u00a0<\/span><\/p>\n<p><span style=\"font-weight: 400;\">Non sottovalutate mai quanto si nasconde nella scatola nera.\u00a0<\/span><\/p>","protected":false},"excerpt":{"rendered":"<p>I ricercatori di Anthropic sono riusciti a identificare milioni di concetti all'interno di Claude Sonnet, uno dei loro LLM avanzati. I modelli di intelligenza artificiale sono spesso considerati scatole nere, cio\u00e8 non \u00e8 possibile \"vedere\" al loro interno per capire esattamente come funzionano.  Quando si fornisce un input a un LLM, questo genera una risposta, ma il ragionamento che sta dietro alle sue scelte non \u00e8 chiaro. L'input entra e l'output esce, e nemmeno gli stessi sviluppatori dell'intelligenza artificiale capiscono cosa succede all'interno di quella \"scatola\".  Le reti neurali creano le proprie rappresentazioni interne delle informazioni quando mappano gli input e gli output durante l'addestramento dei dati. Gli elementi costitutivi di questa<\/p>","protected":false},"author":2,"featured_media":12522,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[84],"tags":[148,105],"class_list":["post-12519","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-industry","tag-anthropic","tag-machine-learning"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v27.4 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>Decoding the AI mind: Anthropic researchers peer inside the &quot;black box&quot; | DailyAI<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/dailyai.com\/it\/2024\/05\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\/\" \/>\n<meta property=\"og:locale\" content=\"it_IT\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Decoding the AI mind: Anthropic researchers peer inside the &quot;black box&quot; | DailyAI\" \/>\n<meta property=\"og:description\" content=\"Anthropic researchers successfully identified millions of concepts within Claude Sonnet, one of their advanced LLMs. AI models are often considered black boxes, meaning you can&#8217;t &#8216;see&#8217; inside them to understand exactly how they work.\u00a0 When you provide an LLM with an input, it generates a response, but the reasoning behind its choices isn\u2019t clear. Your input goes in, and the output comes out \u2013 and even the AI developers themselves don&#8217;t truly understand what happens inside that &#8216;box.&#8217;\u00a0 Neural networks create their own internal representations of information when they map inputs to outputs during data training.\u00a0The building blocks of this\" \/>\n<meta property=\"og:url\" content=\"https:\/\/dailyai.com\/it\/2024\/05\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\/\" \/>\n<meta property=\"og:site_name\" content=\"DailyAI\" \/>\n<meta property=\"article:published_time\" content=\"2024-05-22T18:16:07+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2024-05-24T10:16:44+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/DALL\u00b7E-2024-05-22-19.15.50-Anthropic-researchers-studying-artificial-intelligence-inside-a-high-tech-laboratory-with-a-prominent-black-box-in-the-center.-The-scene-features-adva.webp\" \/>\n\t<meta property=\"og:image:width\" content=\"1792\" \/>\n\t<meta property=\"og:image:height\" content=\"1024\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/webp\" \/>\n<meta name=\"author\" content=\"Sam Jeans\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:creator\" content=\"@DailyAIOfficial\" \/>\n<meta name=\"twitter:site\" content=\"@DailyAIOfficial\" \/>\n<meta name=\"twitter:label1\" content=\"Scritto da\" \/>\n\t<meta name=\"twitter:data1\" content=\"Sam Jeans\" \/>\n\t<meta name=\"twitter:label2\" content=\"Tempo di lettura stimato\" \/>\n\t<meta name=\"twitter:data2\" content=\"6 minuti\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"NewsArticle\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/05\\\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\\\/#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/05\\\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\\\/\"},\"author\":{\"name\":\"Sam Jeans\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#\\\/schema\\\/person\\\/711e81f945549438e8bbc579efdeb3c9\"},\"headline\":\"Decoding the AI mind: Anthropic researchers peer inside the &#8220;black box&#8221;\",\"datePublished\":\"2024-05-22T18:16:07+00:00\",\"dateModified\":\"2024-05-24T10:16:44+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/05\\\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\\\/\"},\"wordCount\":1075,\"publisher\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/#organization\"},\"image\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/05\\\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/05\\\/DALL\u00b7E-2024-05-22-19.15.50-Anthropic-researchers-studying-artificial-intelligence-inside-a-high-tech-laboratory-with-a-prominent-black-box-in-the-center.-The-scene-features-adva.webp\",\"keywords\":[\"Anthropic\",\"machine learning\"],\"articleSection\":[\"Industry\"],\"inLanguage\":\"it-IT\"},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/05\\\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\\\/\",\"url\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/05\\\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\\\/\",\"name\":\"Decoding the AI mind: Anthropic researchers peer inside the \\\"black box\\\" | DailyAI\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/05\\\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/05\\\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/05\\\/DALL\u00b7E-2024-05-22-19.15.50-Anthropic-researchers-studying-artificial-intelligence-inside-a-high-tech-laboratory-with-a-prominent-black-box-in-the-center.-The-scene-features-adva.webp\",\"datePublished\":\"2024-05-22T18:16:07+00:00\",\"dateModified\":\"2024-05-24T10:16:44+00:00\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/05\\\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\\\/#breadcrumb\"},\"inLanguage\":\"it-IT\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/dailyai.com\\\/2024\\\/05\\\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"it-IT\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/05\\\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\\\/#primaryimage\",\"url\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/05\\\/DALL\u00b7E-2024-05-22-19.15.50-Anthropic-researchers-studying-artificial-intelligence-inside-a-high-tech-laboratory-with-a-prominent-black-box-in-the-center.-The-scene-features-adva.webp\",\"contentUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2024\\\/05\\\/DALL\u00b7E-2024-05-22-19.15.50-Anthropic-researchers-studying-artificial-intelligence-inside-a-high-tech-laboratory-with-a-prominent-black-box-in-the-center.-The-scene-features-adva.webp\",\"width\":1792,\"height\":1024},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/2024\\\/05\\\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/dailyai.com\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Decoding the AI mind: Anthropic researchers peer inside the &#8220;black box&#8221;\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#website\",\"url\":\"https:\\\/\\\/dailyai.com\\\/\",\"name\":\"DailyAI\",\"description\":\"Your Daily Dose of AI News\",\"publisher\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/dailyai.com\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"it-IT\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#organization\",\"name\":\"DailyAI\",\"url\":\"https:\\\/\\\/dailyai.com\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"it-IT\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2023\\\/06\\\/Daily-Ai_TL_colour.png\",\"contentUrl\":\"https:\\\/\\\/dailyai.com\\\/wp-content\\\/uploads\\\/2023\\\/06\\\/Daily-Ai_TL_colour.png\",\"width\":4501,\"height\":934,\"caption\":\"DailyAI\"},\"image\":{\"@id\":\"https:\\\/\\\/dailyai.com\\\/#\\\/schema\\\/logo\\\/image\\\/\"},\"sameAs\":[\"https:\\\/\\\/x.com\\\/DailyAIOfficial\",\"https:\\\/\\\/www.linkedin.com\\\/company\\\/dailyaiofficial\\\/\",\"https:\\\/\\\/www.youtube.com\\\/@DailyAIOfficial\"]},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/dailyai.com\\\/#\\\/schema\\\/person\\\/711e81f945549438e8bbc579efdeb3c9\",\"name\":\"Sam Jeans\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"it-IT\",\"@id\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/a24a4a8f8e2a1a275b7491dc9c9f032c401eabf23c3206da4628dc84b6dac5c8?s=96&d=robohash&r=g\",\"url\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/a24a4a8f8e2a1a275b7491dc9c9f032c401eabf23c3206da4628dc84b6dac5c8?s=96&d=robohash&r=g\",\"contentUrl\":\"https:\\\/\\\/secure.gravatar.com\\\/avatar\\\/a24a4a8f8e2a1a275b7491dc9c9f032c401eabf23c3206da4628dc84b6dac5c8?s=96&d=robohash&r=g\",\"caption\":\"Sam Jeans\"},\"description\":\"Sam is a science and technology writer who has worked in various AI startups. When he\u2019s not writing, he can be found reading medical journals or digging through boxes of vinyl records.\",\"sameAs\":[\"https:\\\/\\\/www.linkedin.com\\\/in\\\/sam-jeans-6746b9142\\\/\"],\"url\":\"https:\\\/\\\/dailyai.com\\\/it\\\/author\\\/samjeans\\\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Decodificare la mente dell'intelligenza artificiale: Ricercatori di Anthropic scrutano all'interno della \"scatola nera\" | DailyAI","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/dailyai.com\/it\/2024\/05\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\/","og_locale":"it_IT","og_type":"article","og_title":"Decoding the AI mind: Anthropic researchers peer inside the \"black box\" | DailyAI","og_description":"Anthropic researchers successfully identified millions of concepts within Claude Sonnet, one of their advanced LLMs. AI models are often considered black boxes, meaning you can&#8217;t &#8216;see&#8217; inside them to understand exactly how they work.\u00a0 When you provide an LLM with an input, it generates a response, but the reasoning behind its choices isn\u2019t clear. Your input goes in, and the output comes out \u2013 and even the AI developers themselves don&#8217;t truly understand what happens inside that &#8216;box.&#8217;\u00a0 Neural networks create their own internal representations of information when they map inputs to outputs during data training.\u00a0The building blocks of this","og_url":"https:\/\/dailyai.com\/it\/2024\/05\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\/","og_site_name":"DailyAI","article_published_time":"2024-05-22T18:16:07+00:00","article_modified_time":"2024-05-24T10:16:44+00:00","og_image":[{"width":1792,"height":1024,"url":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/DALL\u00b7E-2024-05-22-19.15.50-Anthropic-researchers-studying-artificial-intelligence-inside-a-high-tech-laboratory-with-a-prominent-black-box-in-the-center.-The-scene-features-adva.webp","type":"image\/webp"}],"author":"Sam Jeans","twitter_card":"summary_large_image","twitter_creator":"@DailyAIOfficial","twitter_site":"@DailyAIOfficial","twitter_misc":{"Scritto da":"Sam Jeans","Tempo di lettura stimato":"6 minuti"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"NewsArticle","@id":"https:\/\/dailyai.com\/2024\/05\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\/#article","isPartOf":{"@id":"https:\/\/dailyai.com\/2024\/05\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\/"},"author":{"name":"Sam Jeans","@id":"https:\/\/dailyai.com\/#\/schema\/person\/711e81f945549438e8bbc579efdeb3c9"},"headline":"Decoding the AI mind: Anthropic researchers peer inside the &#8220;black box&#8221;","datePublished":"2024-05-22T18:16:07+00:00","dateModified":"2024-05-24T10:16:44+00:00","mainEntityOfPage":{"@id":"https:\/\/dailyai.com\/2024\/05\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\/"},"wordCount":1075,"publisher":{"@id":"https:\/\/dailyai.com\/#organization"},"image":{"@id":"https:\/\/dailyai.com\/2024\/05\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\/#primaryimage"},"thumbnailUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/DALL\u00b7E-2024-05-22-19.15.50-Anthropic-researchers-studying-artificial-intelligence-inside-a-high-tech-laboratory-with-a-prominent-black-box-in-the-center.-The-scene-features-adva.webp","keywords":["Anthropic","machine learning"],"articleSection":["Industry"],"inLanguage":"it-IT"},{"@type":"WebPage","@id":"https:\/\/dailyai.com\/2024\/05\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\/","url":"https:\/\/dailyai.com\/2024\/05\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\/","name":"Decodificare la mente dell'intelligenza artificiale: Ricercatori di Anthropic scrutano all'interno della \"scatola nera\" | DailyAI","isPartOf":{"@id":"https:\/\/dailyai.com\/#website"},"primaryImageOfPage":{"@id":"https:\/\/dailyai.com\/2024\/05\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\/#primaryimage"},"image":{"@id":"https:\/\/dailyai.com\/2024\/05\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\/#primaryimage"},"thumbnailUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/DALL\u00b7E-2024-05-22-19.15.50-Anthropic-researchers-studying-artificial-intelligence-inside-a-high-tech-laboratory-with-a-prominent-black-box-in-the-center.-The-scene-features-adva.webp","datePublished":"2024-05-22T18:16:07+00:00","dateModified":"2024-05-24T10:16:44+00:00","breadcrumb":{"@id":"https:\/\/dailyai.com\/2024\/05\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\/#breadcrumb"},"inLanguage":"it-IT","potentialAction":[{"@type":"ReadAction","target":["https:\/\/dailyai.com\/2024\/05\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\/"]}]},{"@type":"ImageObject","inLanguage":"it-IT","@id":"https:\/\/dailyai.com\/2024\/05\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\/#primaryimage","url":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/DALL\u00b7E-2024-05-22-19.15.50-Anthropic-researchers-studying-artificial-intelligence-inside-a-high-tech-laboratory-with-a-prominent-black-box-in-the-center.-The-scene-features-adva.webp","contentUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2024\/05\/DALL\u00b7E-2024-05-22-19.15.50-Anthropic-researchers-studying-artificial-intelligence-inside-a-high-tech-laboratory-with-a-prominent-black-box-in-the-center.-The-scene-features-adva.webp","width":1792,"height":1024},{"@type":"BreadcrumbList","@id":"https:\/\/dailyai.com\/2024\/05\/decoding-the-ai-mind-anthropic-researchers-peer-inside-the-black-box\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/dailyai.com\/"},{"@type":"ListItem","position":2,"name":"Decoding the AI mind: Anthropic researchers peer inside the &#8220;black box&#8221;"}]},{"@type":"WebSite","@id":"https:\/\/dailyai.com\/#website","url":"https:\/\/dailyai.com\/","name":"DailyAI","description":"La vostra dose quotidiana di notizie sull'intelligenza artificiale","publisher":{"@id":"https:\/\/dailyai.com\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/dailyai.com\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"it-IT"},{"@type":"Organization","@id":"https:\/\/dailyai.com\/#organization","name":"DailyAI","url":"https:\/\/dailyai.com\/","logo":{"@type":"ImageObject","inLanguage":"it-IT","@id":"https:\/\/dailyai.com\/#\/schema\/logo\/image\/","url":"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/06\/Daily-Ai_TL_colour.png","contentUrl":"https:\/\/dailyai.com\/wp-content\/uploads\/2023\/06\/Daily-Ai_TL_colour.png","width":4501,"height":934,"caption":"DailyAI"},"image":{"@id":"https:\/\/dailyai.com\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/x.com\/DailyAIOfficial","https:\/\/www.linkedin.com\/company\/dailyaiofficial\/","https:\/\/www.youtube.com\/@DailyAIOfficial"]},{"@type":"Person","@id":"https:\/\/dailyai.com\/#\/schema\/person\/711e81f945549438e8bbc579efdeb3c9","name":"Sam Jeans","image":{"@type":"ImageObject","inLanguage":"it-IT","@id":"https:\/\/secure.gravatar.com\/avatar\/a24a4a8f8e2a1a275b7491dc9c9f032c401eabf23c3206da4628dc84b6dac5c8?s=96&d=robohash&r=g","url":"https:\/\/secure.gravatar.com\/avatar\/a24a4a8f8e2a1a275b7491dc9c9f032c401eabf23c3206da4628dc84b6dac5c8?s=96&d=robohash&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/a24a4a8f8e2a1a275b7491dc9c9f032c401eabf23c3206da4628dc84b6dac5c8?s=96&d=robohash&r=g","caption":"Sam Jeans"},"description":"Sam \u00e8 uno scrittore di scienza e tecnologia che ha lavorato in diverse startup di intelligenza artificiale. Quando non scrive, lo si pu\u00f2 trovare a leggere riviste mediche o a scavare tra scatole di dischi in vinile.","sameAs":["https:\/\/www.linkedin.com\/in\/sam-jeans-6746b9142\/"],"url":"https:\/\/dailyai.com\/it\/author\/samjeans\/"}]}},"_links":{"self":[{"href":"https:\/\/dailyai.com\/it\/wp-json\/wp\/v2\/posts\/12519","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/dailyai.com\/it\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/dailyai.com\/it\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/dailyai.com\/it\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/dailyai.com\/it\/wp-json\/wp\/v2\/comments?post=12519"}],"version-history":[{"count":11,"href":"https:\/\/dailyai.com\/it\/wp-json\/wp\/v2\/posts\/12519\/revisions"}],"predecessor-version":[{"id":12573,"href":"https:\/\/dailyai.com\/it\/wp-json\/wp\/v2\/posts\/12519\/revisions\/12573"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/dailyai.com\/it\/wp-json\/wp\/v2\/media\/12522"}],"wp:attachment":[{"href":"https:\/\/dailyai.com\/it\/wp-json\/wp\/v2\/media?parent=12519"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/dailyai.com\/it\/wp-json\/wp\/v2\/categories?post=12519"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/dailyai.com\/it\/wp-json\/wp\/v2\/tags?post=12519"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}