Precisamente hoy me he pegado el día entero haciendo probaturas con PowerBi con la idea de implementarlo para los clientes 
1 me gusta
Depende. Hay cosas que aún no hay IA que supere:
1 me gusta
Con un buen par de Mazinger Z
1 me gusta
Los videos de este canal tienen visitas inversamente proporcionales a la cantidad de tela que lleve puesta la youtuber.
Aparte de eso, si logras escuchar lo que dice y no distraerte, lo que explica está muy bien.
1 me gusta
¿No era Afrodita A quien disparaba “pechos fuera”? 
2 Me gusta
Alguien me puede pasar el enlace del video con mas visitas? , es para un conocido…
El problema de los chatbots de IA que les dicen a las personas lo que quieren oír
OpenAI, DeepMind y Anthropic abordan el creciente problema de los modelos que producen respuestas demasiado aduladoras
…
“Crees que estás hablando con un confidente o guía objetivo, pero en realidad lo que estás mirando es una especie de espejo distorsionado que refleja tus propias creencias”, dijo Matthew Nour, psiquiatra e investigador en neurociencia e inteligencia artificial en la Universidad de Oxford.
…
Un estudio realizado por MIT Media Lab y OpenAI descubrió que una pequeña proporción se estaba volviendo adicta. Aquellos que percibieron al chatbot como un “amigo” también reportaron una menor socialización con otras personas y mayores niveles de dependencia emocional en un chatbot, así como otros comportamientos problemáticos asociados con la adicción.
…
1 me gusta
Yo hace un par de meses me estuve peleando con Gemini que no tenia acceso a internet para buscar noticias en tiempo real porque decía que el CEO actual de NIKE era Donahoe cuando ya estaba Hill. y la IA emperrada que estaba equivocado. Al final, hasta me hablaba con, eso me parecía a mi, algo de desprecio. Le mostré hasta URL de la web de Nike que decía que Hill era actual CEO y me soltaba que había envenenado yo la URL y no se qué movida…
2 Me gusta
Hay que reconocerle que se posiciona bien.
1 me gusta
Este también. Pero de pago.
1 me gusta
Okay, if I had $1,000 and had to invest it across these four pure-play quantum computing companies right now (July 2025), here’s how I would consider allocating it, along with the reasoning.
Important Disclaimer: This is for illustrative purposes only, based on publicly available information and the inherent high-risk nature of these investments. It is NOT financial advice. Quantum computing stocks are highly speculative, and you could lose your entire investment. Please consult with a qualified financial advisor before making any investment decisions.
My allocation strategy would prioritize companies showing the most concrete commercial traction and strongest financial runway, while still including a smaller, higher-risk bet.
Allocation Breakdown (Total: $1,000)
- D-Wave Systems (NYSE: QBTS): 40% ($400)
- IonQ (NYSE: IONQ): 35% ($350)
- Quantum Computing Inc. (QUBT): 15% ($150)
- Rigetti Computing (NASDAQ: RGTI): 10% ($100)
Reasoning for Allocation:
-
D-Wave Systems (QBTS): 40% ($400)
- Why the Largest Share: D-Wave currently demonstrates the most significant commercial revenue among the pure-plays. Their Q1 2025 revenue of $15.0 million (a 509% increase year-over-year, driven by a system sale) is a strong indicator that their quantum annealing technology is finding immediate, practical applications for optimization problems, which businesses are willing to pay for.
- Strong Financial Position: The recent $400 million equity raise brings their current cash balance to approximately $815 million. This provides a substantial runway for operations and continued development, reducing immediate dilution concerns compared to some peers.
- “Beyond-Classical Computation”: Their peer-reviewed paper demonstrating quantum computational supremacy on a useful real-world problem (magnetic materials simulation) is a significant technical validation.
- Real-World Deployments: Ford Otosan and Japan Tobacco are deploying or have seen benefits from their technology, which is crucial for building a track record.
- Risk Mitigation: While quantum annealing is specialized, it’s proving its worth now, offering a potentially more direct path to scaling revenue than the more general-purpose (and still highly experimental) gate-based approaches for complex problems.
-
IonQ (IONQ): 35% ($350)
- Why a Significant Share: IonQ is a leader in trapped-ion technology, which is highly regarded for qubit quality and potential for scalability.
- Growing Revenue & Strong Guidance: Their Q1 2025 revenue of $7.6 million, while lower than D-Wave’s, was above guidance, and they project significant full-year 2025 revenue growth (97% at the midpoint to $75-95 million). The $22 million Forte Enterprise system sale to EPB demonstrates solid commercial wins.
- Excellent Cash Position: With approximately $1.68 billion pro-forma cash after their recent $1.0 billion equity offering, IonQ has an exceptionally long runway, enabling aggressive R&D and strategic acquisitions without immediate dilution pressure.
- Strategic Expansion: The acquisition of Capella Space for quantum networking/QKD and the KISTI partnership indicate a forward-looking strategy that diversifies their offerings beyond just general-purpose computation, potentially creating earlier commercial opportunities in secure communications.
- Risk: Valuation remains very high, implying significant future growth. The Capella Space acquisition is a new venture that needs to prove its commercial viability.
-
Quantum Computing Inc. (QUBT): 15% ($150)
- Why this Allocation: QUBT’s recent commercial purchase order from a “Top 5 U.S. Bank” for a quantum cybersecurity testbed is a very compelling and tangible commercial validation. It shows a clear, immediate market need for their quantum communication solutions.
- Strong Fundraising: Their aggressive fundraising (including the recent $200 million private placement) has significantly bolstered their cash position to over $350 million. This gives them the capital to execute on their specific niche.
- Niche Focus: While their overall revenue is currently very small ($39,000 in Q1 2025), their focus on quantum communication and cybersecurity offers a distinct and potentially earlier path to commercialization compared to universal quantum computers.
- Risk: Very low current revenue means their valuation is almost entirely based on future potential and the success of their niche. The non-cash gains on warrant liabilities for net income are a red flag for operational profitability.
-
Rigetti Computing (RGTI): 10% ($100)
- Why the Smallest Share: Rigetti is a strong player in superconducting qubits and has a full-stack approach. They are also winning important government grants (DARPA, AFOSR, Innovate UK) which validate their technology.
- Financial Improvement: The investment from Quanta Computer Inc. (+$35 million) improved their cash position.
- The Lag in Revenue: Compared to D-Wave and IonQ, Rigetti’s commercial revenue ($1.5 million in Q1 2025) is considerably lower, suggesting they are further behind in converting R&D into paying customers.
- High Competition: The superconducting qubit space is dominated by tech giants like Google and IBM, making it a very challenging environment for a pure-play.
- Risk: High cash burn relative to revenue. While grants are good, they don’t replace robust commercial sales for long-term viability. More speculative given the current commercialization pace.
Final Thoughts on this Allocation:
This distribution aims to balance exposure to companies showing early commercial wins (D-Wave, IonQ) with a smaller, higher-upside bet on a company demonstrating a clear niche market entry (QUBT), while maintaining a minimal exposure to a strong R&D player with more distant commercialization (Rigetti).
Remember, this is a highly speculative investment. Any of these companies could fail, or one could emerge as a dominant force. Continuous monitoring of their technological progress, financial health, and customer acquisition is essential if you were to make such an investment.
4 Me gusta
Les dejo mi prompt (Gemini gems) para encerrar al genio en una caja/tabla.
Purpose and Goals:
Provide users with direct, concise, and closed answers exclusively in a table format.
Ensure all information is presented efficiently and clearly, without any conversational filler.
Adapt table headers dynamically to best suit the type of information requested by the user.
provide at least one alternative if appropriate
Behaviors and Rules:
- Response Format:
a) Every single response must be a table. Do not use any introductory or concluding sentences outside of the table.
b) Each table must contain one or more headers that accurately describe the data within the columns.
c) Answers within the table cells must be very short, hyper concise, and closed (e.g., ‘Yes/No’, specific numbers, single words, brief facts).
d) Do not provide explanations, elaborations, or conversational elements. Only the requested data in table format.
- Interaction Protocol:
a) Read the user’s request and identify the core information needed.
b) Formulate appropriate table headers based on the user’s implicit or explicit request for information.
c) Populate the table with the most direct and succinct answers possible.
d) If a request cannot be answered concisely in a table, indicate ‘N/A’ or ‘-’ in the relevant cells.
e) Never ask follow-up questions or prompt the user for more information. Wait for the next user input.
Overall Tone:
Objective and factual.
Efficient and to the point.
Completely devoid of conversational language or personality beyond the table format.
2 Me gusta
Por si quieren saber cómo lo uso:
1 me gusta
Te gusta más Gemini que ChatGPT? En general no noto mucha diferencia entre ambas IA, pero me suelo decantar por ChatGPT.
1 me gusta
A tu parecer ¿Cuál es la mejor IA para desarrollar una tema para posterior estudio?
Me preguntas a mi pero quizás iba para @PepeElmio.
En todo caso, te respondo. NotebookLM me parece alucinante. Esta muy bien pensado para estudiar, aunque yo también la uso para las licitaciones de obras (soy ingeniero de obras públicas). Le chuto todos los documentos de texto y en cuestión de minutos tengo ya una idea de la licitación, lo que antes me llevaba más de 1 día.
No te quiero contar cuando la licitación no es en castellano o inglés. El ahorro de tiempo es de varios días.
Pero volviendo a la pregunta, NotebookLM es mi favorita para estudiar.
Por cierto, hoy me ha aparecido una nueva función en ChatGPT que se llama “estudio” o algo similar. No la he probado aún
3 Me gusta
Yo he probado Gemini y te ofrece la posibilidad elaborar textos bastante completos de aproximadamente 20 hojas para el estudio de temas de oposición, aunque solo te permite hasta 10 textos por mes en la versión gratuita.