knowledge connexions

The state of AI in 2020: Democratization, industrialization, and the way to artificial general intelligence

From fit for purpose development to pie in the sky research, this is what AI looks like in 2020.

After releasing what may well have been the most comprehensive report on the State of AI in 2019, Air Street Capital and RAAIS founder Nathan Benaich and AI angel investor and UCL IIPP visiting professor Ian Hogarth are back for more.

In the State of AI Report 2020, Benaich and Hogarth outdid themselves. While the structure and themes of the report remain mostly intact, its size has grown by nearly 30%. This is a lot, especially considering their 2019 AI report was already a 136 slide long journey on all things AI.

The State of AI Report 2020 is 177 slides long, and it covers technology breakthroughs and their capabilities, supply, demand, and concentration of talent working in the field, large platforms, financing, and areas of application for AI-driven innovation today and tomorrow, special sections on the politics of AI, and predictions for AI.

ZDNet caught up with Benaich and Hogarth to discuss their findings.

AI democratization and industrialization: Open code and MLOps

We set out by discussing the rationale for such a substantial contribution, which Benaich and Hogarth admitted to having taken up an extensive amount of their time. They mentioned their feeling is that their combined industry, research, investment, and policy background and currently held positions give them a unique vantage point. Producing this report is their way of connecting the dots and giving something of value back to the AI ecosystem at large.

Coincidentally, Gartner’s 2020 Hype cycle for AI was also released a couple of days back. Gartner identifies what it calls 2 megatrends that dominate the AI landscape in 2020: Democratization and industrialization. Some of Benaich and Hogarth’s findings were about the massive cost of training AI models, and the limited availability of research. This seems to contradict Gartner’s position, or at least imply a different definition of democratization.

Benaich noted that there are different ways to look at democratization. One of them is the degree to which AI research is open and reproducible. As the duo’s findings show, it is not: only 15% of AI research papers publish their code, and that has not changed much since 2016.

Hogarth added that traditionally AI as an academic field has had an open ethos, but the ongoing industry adoption is changing that. Companies are recruiting more and more researchers (another theme the report covers), and there is a clash of cultures going on as companies want to retain their IP. Notable organizations criticized for not publishing code include OpenAI and DeepMind:

“There’s only so close you can get without a sort of major backlash. But at the same time, I think that data clearly indicates that they’re certainly finding ways to be close when it’s convenient,” said Hogarth.


Industrialization of AI is under way, as open source MLOps tools help bring models to production

As far as industrialization goes, Benaich and Hogarth pointed towards their findings in terms of MLOps. MLOps, short for machine learning operations, is the equivalent of DevOps for ML models: Taking them from development to production, and managing their lifecycle in terms of improvements, fixes, redeployments, and so on.

Some of the more popular and fastest-growing Github projects in 2020 are related to MLOps, the duo pointed out. Hogarth also added that for startup founders, for example, it’s probably easier to get started with AI today than it was a few years ago, in terms of tool availability and infrastructure maturity. But there is a difference when it comes to training models like GPT3:

“If you wanted to start a sort of AGI research company today, the bar is probably higher in terms of the compute requirements. Particularly if you believe in the scale hypothesis, the idea of taking approaches like GPT3 and continuing to scale them up. That’s going to be more and more expensive and less and less accessible to new entrants without large amounts of capital.

The other thing that organizations with very large amounts of capital can do is run lots of experiments and iterates in large experiments without having to worry too much about the cost of training. So there’s a degree to which you can be more experimental with these large models if you have more capital.

Obviously, that slightly biases you towards these almost brute force approaches of just applying more scale, capital and data to the problem. But I think that if you buy the scaling hypothesis, then that’s a fertile area of progress that shouldn’t be dismissed just because it doesn’t have deep intellectual insights at the heart of it.”

How to compete in AI

This is another key finding of the report: huge models, large companies, and massive training costs dominate the hottest area of AI today: NLP (Natural Language Processing). Based on variables released by Google et. al., research has estimated the cost of training NLP models at about $1 per 1000 parameters.

That means that a model such as OpenAI’s GPT3, which has been hailed as the latest and greatest achievement in AI, could have cost tens of millions to train. Experts suggest the likely budget was $10 million. That clearly shows that not everyone can aspire to produce something like GPT3. The question is: Is there another way? Benaich and Hogarth think so and have an example to showcase.

PolyAI is a London-based company active in voice assistants. They produced and open-sourced a conversational AI model (technically, a pre-trained contextual re-ranker based on transformers) that outperforms Google’s BERT model in conversational applications. PolyAI’s model not only performs much better than Google’s, but it required a fraction of the parameters to train, meaning also a fraction of the cost.


PolyAI managed to produce a machine learning language models that performs better than Google in a specific domain, at a fraction of the complexity and cost.

The obvious question is: How did PolyAI do it? This could be an inspiration for others, too. Benaich noted that the task of detecting intent and understanding what somebody on the phone is trying to accomplish by calling is solved in a much better way by treating this problem as what is called a contextual re-ranking problem:

“That is, given a kind of menu of potential options that a caller is trying to possibly accomplish based on our understanding of that domain, we can design a more appropriate model that can better learn customer intent from data than just trying to take a general purpose model — in this case BERT.

BERT can do OK in various conversational applications, but just doesn’t have kind of engineering guardrails or engineering nuances that can make it robust in a real world domain. To get models to work in production, you actually have to do more engineering than you have to do research. And almost by definition, engineering is not interesting to the majority of researchers.”


SoftwareApplication – Test

Test 1

Test 1

Test 2

Test 2


Si tu empresa recibe una exorbitante cantidad de mensajes por día, necesitas

Hace poco hablaba con un amigo cercano que tiene una clínica, donde en promedio se realizan más de 800 cirugías por mes, mi amigo preocupado me comentaba que más de 6000 clientes aproximadamente escriben solicitando información pero que no llegaban a atenderlos a todos, y era tanto el colapso en su sistema de atención que muchas veces el 50% de los clientes al parecer no quedaban  satisfechos con las respuestas brindadas por parte de su equipo de servicio al cliente. Él se preguntaba, ¿Qué debo hacer?, ¿Cuántas personas tendría que contratar para dar respuesta a la misma pregunta de casi todos los usuarios? ¿Tal vez 4 o 7 personas? Solo para responder la misma pregunta una y otra vez.

Adicional a eso se encuentro con una gran cantidad de información de potenciales clientes que su equipo de trabajo no tenia manera de controlar, ya que tendrían que dedicar horas y horas a revisar cada mensaje de cada celular Â¿Cómo hacer que tu empresa sea más productiva cuando se tiene más de 5000 mensajes por responder? En este artículo te ayudaremos a descubrirlo.

El mundo tecnológico avanza cada vez más rápido y las empresas siempre siguen a los clientes hacia nuevas plataformas. Actualmente podría decirse que no hay una plataforma más grande y prominente que WhatsApp en lo que se refiere a contacto con clientes. El elegante estilo y el diseño lo han convertido en un éxito para casi todos los grupos demográficos, independientemente de su edad, género, ubicación o idioma.

Desde su creación, WhatsApp ha revolucionado la forma en que nos comunicamos. Desde entonces se han creado muchas aplicaciones de chat, pero ninguna con el mismo éxito o alcance de usuarios de WhatsApp

No es de extrañar que el mundo empresarial haya estado ansioso por atraer a los clientes con un medio global, fácil de usar y muy productivo.

Cuando anunciaron que llegaba la solución WhatsApp Business API el mundo empresarial lo celebro, ya que esta nueva versión de la plataforma traería todo el potencial de aprovechar las comunicaciones omnicanal, respaldar el auge del comercio conversacional y brindar personalización a las conversaciones con los clientes. Proveedores de soluciones tecnológicas de todo el mundo se volcaron a usar esta herramienta que ayuda a todas las empresas a posicionarse frente a su cliente

Pero, ¿Por qué causo todo este alboroto WhatsApp Business API?

En primer lugar, está la demanda de los clientes. De acuerdo con el reporte anual 2019-2020 de Burst SMS los clientes quieren la opción de enviar mensajes de texto a las empresas, porque los mensajes de texto se convirtieron en la forma principal en que nos comunicamos. La mensajería SMS satisface parcialmente esta necesidad, pero carece de la facilidad de conversación y las funcionalidades multimedia a las que estamos acostumbrados. El hecho es que las aplicaciones de chat se han integrado a la perfección en nuestros estilos de vida y ahora simplemente no podemos dejar de usarlas.

En segundo lugar, desde un punto de vista empresarial, existe una necesidad real de modernizar las comunicaciones con los clientes. La forma en que las empresas se comunican con los clientes debe reflejar cómo ha cambiado el mundo desde el comienzo de los mensajes de texto. Ahora, los clientes exigen respuestas personalizadas, comunicaciones en tiempo real y gratificación instantánea. Las interacciones omnicanal se han convertido en la regla, no en la excepción.

El dejar en espera para hablar con un representante de servicio al cliente está creciendo cada vez más como una práctica inaceptable. â›”

➤ WhatsApp Business API: Un potente canal de mensajería empresarial.

WhatsApp Business API tiene el potencial de cambiar el juego. Dado que WhatsApp viene ya preinstalado en muchos teléfonos inteligentes en todo el mundo, la adopción de usuarios ha crecido exponencialmente. Eso significa solo una cosa, que sus clientes probablemente ya estén usando la aplicación. Para las empresas, esto representa una inmensa oportunidad de crear conexiones de marca con sus clientes, y mucho más.

El uso de WhatsApp Business API ayuda a las empresas a involucrarse mucho más con los clientes desde notificaciones unidireccionales como recordatorios de citas, alertas de envío, notificaciones de pago, códigos de verificación y tarjetas de embarque, hasta encuestas y conversaciones bidireccionales de atención al cliente. Con la nueva API, las empresas pueden aprovechar la inmediatez y el toque personal, al tiempo que pueden llegar a los usuarios a nivel internacional. Es más, su cifrado y cuentas verificadas brindan seguridad a las empresas y los usuarios en un mundo cada vez más digital. Cuando se usa de manera apropiada, la solución WhatsApp Business API puede ayudar a las empresas a brindar la experiencia omnicanal que los clientes piden con la seguridad que necesitan.

Pero las ventajas de la solución WhatsApp Business API no se detienen ahí. Seguro te preguntarás que otras maravillas trae esta herramienta. Aquí te dejamos algunas:

  • Mejorar velocidad de respuesta (SLA).
  • Ahorro de costos en agentes y Call-Center.
  • Aumento en la satisfacción del cliente a 98%.
  • Cumplir políticas de privacidad de datos (GDPR).
  • Aumento de productividad con asignaciones por equipo o agente.
  • Mejorar la relación y comunicación con tus clientes.
  • Habilita soporte y atención digital 24/7.
  • Seguridad y encriptación de mensajes end-to-end.
  • Habilitar un número oficial con perfil de Empresa verificado. ✅

Hospital – Doctor


Cuore veg di couscous di barbabietola

Salutiamo il mese di Ottobre e diamo il via al periodo Veg, ebbene si con il couscous è possibile realizzare moltissime ricette vegane e questa di oggi è solo la prima: Cuore veg di couscous di barbabietola! La ricetta è della nostra Sara Bonaccorsi, food blogger di Giallo Zafferano!

Ingredienti per il cous cous:

  • 200 gr cous cous alla barbabietola
  • 200 ml acqua
  • 1 cucchiaino sale fino
  • 1 cucchiaio olio evo
  • 1 pizzico di zenzero in polvere

Ingredienti per il condimento:

  • Mezza cipolla rossa
  • 2 cucchiai olio evo
  • 100 gr cannellini lessati
  • 100 gr fagioli neri lessati


Iniziate lessando i cannellini e i fagioli neri, in 2 pentole differenti. Affettate la cipolla, non troppo finemente. Soffriggetela in padella insieme all’olio extra vergine di oliva; unite i cannellini e i fagioli neri, e fateli insaporire a fiamma vivace per qualche minuto.

Reidratate il cous cous: in una padella bassa e larga portate a bollore l’acqua, con il sale e lo zenzero. Togliete dal fuoco, unite il cous cous di barbabietola; coprite con un coperchio e fate riposare per 5 minuti. Condite con un cucchiaio di olio extra vergine di oliva e sgranate benissimo con una forchetta.

Con un taglia biscotti a forma di cuore, realizzate il cuore di cous cous. Condite con i fagioli e le cipolle e servite subito.
Buon cous cous a tutti!

Couscous Bio Integrale

Couscous Medio Bio


About + Contact