⚔️ Sir Demis Hassabis: the newly minted knight of AI

Google DeepMind CEO Hassabis recently received the title of Sir and a knighthood for his contributions to AI development. He is a Fellow of the UK's Royal Academy of Engineering and the Royal Society of London and part of Time magazine’s 100 Most Influential People in AI list.

Boasting a PhD in neurobiology, this visionary scientist believes that understanding the brain's mechanisms will help make incredible breakthroughs in AI.

BIOGRAPHY

Hassabis was born in 1976 in London. His father was Greek Cypriot, and his mother was Chinese Singaporean.

♟️ He learned to play chess at age 4. By 13, this chess prodigy had an Elo rating of 2300 points.

At age 8, Hassabis bought his first computer and taught himself how to program from books.

He entered Cambridge at the age of 16.

🧠 Hassabis did his PhD in cognitive neuroscience in London and co-authored a study on neurobiology and AI conducted by MIT and Harvard.

In 2010, Hassabis founded DeepMind Technologies lab with Shane Legg and Mustafa Suleyman. Elon Musk was among the investors. In 2014, Google acquired DeepMind for $650 million.

🥇 In 2016, DeepMind released AlphaGo, an AI program that beat the top professional player of Go.

The Hassabis team also developed The AlphaFold and AlphaFold2 AI programs capable of predicting the shape of proteins down to atoms and building three-dimensional protein models. This breakthrough will help create new drugs and treat severe diseases like Alzheimer's.

DeepMind merged with Google Brain in 2023 and released the Gemini 1.5 Pro in 2024.

👾 One of Hassabis’ favorite games is a strategy game called The Battle of Polytopia (Elon Musk loves it, too), and his favorite authors include Noah Harari, Isaac Asimov, and Ian Banks.

▶️ Video interview with Lex Friedman

🎤 Interview for The Verge

You might also be interested in:
Mustafa Suleyman on How AI Will Transform the World

@hiaimediaen

🔝 GPT-4-Turbo has taken the top spot in the ranking of LLMs. Let's explain how it's compiled

Last week, OpenAI released an updated version of the GPT-4-Turbo model. According to reviews, it excels at generating code and produces texts with less "fluff." After its release, GPT-4-Turbo claimed the first position in the ranking of large language models (LLMs) — the Arena Leaderboard. This ranking is based on user voting and pairwise comparisons.

❓ What do the parameters in the Arena ranking mean?

⚪️ Arena Elo — the rating of LLMs, similar to the one used in chess. It indicates how well a model performs in competitions against others. In these competitions, users choose the best model. The comparison happens in a blind testing mode: users are presented with two windows to interact with unknown models, and they can converse with them until they determine the best LLM based on their personal preferences.

⚪️ CI, 95% — the confidence interval of the model's rating. What does this mean: for example, if the GPT-4-Turbo model has an Elo rating of 1260 and a CI of 95% of +5/-5; it means that with a 95% probability, the actual rating of the model will be within the range of 1255–1265 (+5 and -5 from the rating of 1260). The CI helps track the objectivity of the LLM's position in the table, which is why three models immediately occupy the 1st place: their ratings "overlap" in their confidence intervals.

⚪️ Votes — the number of user votes in Arena. As of today, Arena operates with 670 thousand votes.

⚪️ Organization — the company that owns the LLM.

⚪️ Knowledge Cutoff — the period the model's data is trained; for example, GPT-4-Turbo has world knowledge until December 2023.

👍 How can I test it?

In our @GPT4Telegrambot, you can work directly with four leading models: GPT-4-Turbo, GPT-4-0125, and Claude 3 — versions of Opus and Sonnet.

📕 More stories:

⚡️ Claude 3: a new AI model from the main competitor of OpenAI

📎 @GPT4Telegrambot Now Features File Recognition

#OpenAI #Claude @hiaimediaen

❓ What is AGI?

Even the major Big Tech companies don't quite agree on where we currently are regarding the level of AI development. In a recent lawsuit against OpenAI, Musk claimed that GPT-4 is early AGI, essentially equal to human intelligence.

If true, then according to OpenAI's mission statement, it should serve the "benefit of humanity" and not just profit Microsoft, which owns an exclusive license to OpenAI's products. Microsoft keeps claiming its rights to GPT-4. The company believes that the AGI has not yet been reached.

🖥 Where does the AGI concept come from?

While the term "Artificial Intelligence" appeared in 1956, the concept of AGI came into use in 2007, after a book by SingularityNET founder Ben Herzl called "Artificial General Intelligence" came out.

In 2010, while addressing potential investors at the Singularity Conference about the newly launched DeepMind Lab, Demis Hassabis was among the first to talk about achieving AGI. This was the startup's goal stated in the company's business plan.

🤔 What does AGI mean now?

Altman defines AGI as the "equivalent of a median human that you could hire as a co-worker." Hassabis says AGI "should be able to do pretty much any cognitive task humans can do."

📝 How does "general AI" differ from "ordinary" AI?

Most AI models still perform narrowly focused tasks. AGI must be able to independently solve many versatile tasks, self-learn quickly, adapt to context, and produce new knowledge.

⏳ When will we achieve AGI?

In the 1960s, scientists predicted the arrival of "thinking machines" as early as the late 1980s. Now, the most cited date is 2031. Still, the head of Meta AI, Yann LeCun, believes that the transition to AGI may take a few decades.

🆘 Is AGI dangerous for humans?

Elon Musk believes AGI is "the greatest existential threat we face today." Yet Demis Hassabis is convinced that with the advent of AGI, humanity will reach a new level of development and come closer to creating Super Intelligence (ASI).

#ElonMusk @hiaimediaen

📣 Hi, everyone! We want to share a significant milestone with you: the number of users of our bot has exceeded 11.5 million people! The number of readers in this channel has reached 100,000 people!

We are always happy to receive your feedback. Write comments, give likes, and suggest new topics. Thank you for being with us 👍

As usual, we are sharing the most interesting news from the world of AI for the 16th week of 2024 in our Sunday digest.

AI NEWS

🤖 GPT-4-Turbo has taken the top spot in the ranking of LLMs. This ranking is based on user voting and pairwise comparisons.

🦙 Meta AI has released Llama 3 — the largest open-source language model. You can test it for free.

👾 VASA — hyper-realistic talking avatar. A new AI model from Microsoft enables the creation of avatars that move and speak like real people.

SAVE THIS — IT'S HELPFUL

🎨 AI for logo creation. If your company needs a logo and the designer is looking for inspiration, Brandmark will help you swiftly develop many options while considering the company's domain.

TO READ

⚔️ Sir Demis Hassabis: the knight of AI. Google DeepMind CEO Hassabis recently received the title of Sir and a knighthood for his contributions to AI development.

🍻 A better beer with a little help … from AI! AI analyzed data about beer and learned to predict its popularity better than humans.

📚 5 new books on the past, present, and future of artificial intelligence from top global scientists and researchers.

🤔 What is AGI? The mission of OpenAI, like that of Google DeepMind, is to create Artificial General Intelligence. What is it and when can we expect it?

#AIweek | @hiaimediaen