Introducing Gemini - our largest and most capable AI model, with huge advances in multimodal capabilities and sophisticated reasoning! The possibilities of a world empowered responsibly by AI are really exciting. Very proud of the whole team!
We’re excited to announce 𝗚𝗲𝗺𝗶𝗻𝗶:
@Google
’s largest and most capable AI model.
Built to be natively multimodal, it can understand and operate across text, code, audio, image and video - and achieves state-of-the-art performance across many tasks. 🧵
Welcome
@Rob_Fergus
to
@DeepMind
. He's a computer vision and deep learning pioneer & a leading scientist in AI. I’m delighted he’s joined
@DeepMind
(remaining Prof @ NYU), looking forward to working with him again many years after we were at NYU.
AI is already enabling progress in fundamental research areas, it also holds great potential to address global challenges like the climate crisis. Read more about recent work from our team to help tackle climate change and build AGI:
Most real world problems are complex and have unknown dynamics, so we need agents that can learn to plan on their own. MuZero is a very exciting development towards that goal - congrats to the team!
In 2016, AlphaGo was introduced. Two years later, its successor - AlphaZero - showed significant progress in Go, chess and shogi. Today in
@Nature
, our team describes MuZero, a significant step forward in the pursuit of general-purpose algorithms:
A very exciting step forward in nuclear fusion with Deep RL, in collaboration with
@EPFL
. This is a great example of how AI research is making impact on one of the most important problems in the world. Congrats to the whole team!
Today in
@nature
, with
@EPFL
, the first deep reinforcement learning system that can keep nuclear fusion plasma stable inside its tokamaks, opening new avenues to advance nuclear fusion research.
Paper:
Great work from the team. AlphaCode is a great example of problem solving by AI that requires a combination of creativity, critical thinking and advanced coding.
This is only the beginning - there’s much more to discover and we can’t wait to continue this exploration.
Introducing
#AlphaCode
: a system that can compete at average human level in competitive coding competitions like
@codeforces
. An exciting leap in AI problem-solving capabilities, combining many advances in machine learning!
Read more: 1/
Today is a very important day for research and a testament to the enormous amount of work done by the team. As researchers pursue solutions to significant problems facing humankind, we hope they will benefit from these new insights - congratulations to all involved!
Today with
@emblebi
, we're launching the
#AlphaFold
Protein Structure Database, which offers the most complete and accurate picture of the human proteome, doubling humanity’s accumulated knowledge of high-accuracy human protein structures - for free: 1/
Looking forward to working with many brilliant research colleagues from across the
@DeepMind
&
@Google
Research Brain. Excited to join forces to accelerate progress towards a world where
#AI
helps solve humanity’s biggest challenges
The phenomenal teams from Google Research’s Brain and
@DeepMind
have made many of the seminal research advances that underpin modern AI, from Deep RL to Transformers. Now we’re joining forces as a single unit, Google DeepMind, which I’m thrilled to lead!
Congrats to Rich Sutton, Doina Precup and Satinder Singh for winning the 2019 Classical AIJ Paper Award with their seminal 1999 paper on temporal abstractions!
@DeepMindAI
@McGillU
@UAlberta
@Umich
We’ve posted an updated version of our SARS-CoV-2 structure predictions, and discuss how well our prediction of ORF3a matched experimental data recently released by
@kerndm
and the Brohawn lab
@berkeleyMCB
.
Read more in the updated post:
Congratulations to everyone at
#CASP14
today. This level of accuracy reinforces our hope that AI could become the key for understanding the world around us.
#AlphaFold
We’re excited about the potential impact
#AlphaFold
may have on the future of biological research and scientific discovery. Thank you to the CASP organisers & the whole community - we look forward to the many years of hard work and discovery ahead:
Excited to share our recent work on Language Models. They draw on the expertise of our multidisciplinary teams - including Language, Deep Learning, Ethics, and Safety. Great job to everyone who contributed! Read more on the blog.
Today we're releasing three new papers on large language models. This work offers a foundation for our future language research, especially in areas that will have a bearing on how models are evaluated and deployed: 1/
Last year we presented
#AlphaFold
v2 which predicts 3D structures of proteins down to atomic accuracy. Today we’re proud to share the methods in
@Nature
w/open source code. Excited to see the research this enables. More very soon!
Global representation is key to building a stronger and more inclusive AI industry. I’m very proud that the DeepMind scholarship programme has grown to include universities across the world from many regions that have been historically underrepresented in our field.
Our scholarship programme was established in an effort to help build a stronger and more inclusive AI community. This week, we welcomed our largest and most geographically diverse cohort yet. (1/3)
Great work from the team! Sparrow, our new dialogue agent using human feedback to maximise helpfulness, correctness, and harmlessness. We hope Sparrow will be a significant step forward in safer and more useful dialogue agents.
Large language models can exhibit falsehoods, discriminatory language, and other unsafe behaviour. Introducing Sparrow: a dialogue agent that can search the internet and is trained to be more helpful, correct, and harmless using RL from human feedback: 1/
Two of our recent projects published in
@Nature
today! Our work on protein folding
#AlphaFold
and another exploring the link between distributional RL and dopamine in the brain. Huge congrats to everyone who contributed to make these scientific achievements possible!
@DeepMindAI
Exciting to see our relationship with ITU grow through the DeepMind scholarship programme! We hope this helps encourage students to consider postgraduate study in AI.
Thrilled to announce another global addition to the DeepMind scholarship programme with
@itu_ai_research
!
Read more about the qualifications, benefits, application process and timeline today:
Seems to solidly compete with GPT-4 on benchmarks.
Google has existing customers and surfaces to start the feedback loop, without worrying about adoption.
And Google will use TPUs for inference, so doesn't have to pay NVIDIA their 70% margins (like
@OpenAI
and
@Microsoft
has to
Amazing to see the progress of WaveNet: first w/
@Google
’s Assistant & now helping ALS advocate
@TShawsTruth
recreate & hear his original voice as part of
@Google
Project Euphonia. More in
@Youtube
's
#AgeOfAI
documentary
[1/2] Using WaveNet technology to reunite speech-impaired users with their original voices: in collaboration with
@GoogleAI
, we've applied new iterations including WaveRNN to help ALS advocate
@TShawsTruth
recreate & hear his original speaking voice
Blog:
Delighted and honoured to receive a Knighthood for services to AI. It’s been an incredible journey so far building
@GoogleDeepMind
over the past 15 years, helping accelerate the field and grow the UK & global AI ecosystems. Thanks to everyone who helped make this dream possible!
Open sourcing bsuite. Automated evaluation and analysis of agents on RL benchmarks. We hope this will help reproducible and accessible research on core problems in RL. Looking forward to seeing results from the community!
We are excited to release Behaviour Suite for Reinforcement Learning, or ‘bsuite’ – a collection of carefully-designed experiments that investigate core capabilities of RL agents
GitHub:
Paper:
Congrats to both our teams for these very exciting works at
#ICLR2021
and thanks to the award committee for recognising. A new approach to PCA as a competitive game and learned physics simulation on irregular meshes with graph neural networks. Looking forward to the conference!
Huge congratulations to John Jumper for being included in the
@TIME
100 Next list for his tremendous contribution to
#AlphaFold
and its potential to accelerate biological research. Very well deserved!
#Time100Next
I couldn't agree more with Lila - we all have a responsibility to show our support for the Black community and combat racism within our own spaces.
#BlackLivesMatter
Excited to see our very own David Silver as the recipient of the 2019
@TheOfficialACM
Prize in Computing. His contribution has inspired countless innovations within the field of deep reinforcement learning - a huge achievement. Congrats Dave!
Congratulations to David Silver, recipient of this year’s
@TheOfficialACM
prize in Computing! David is being recognised for his role developing
#AlphaGo
and
#AlphaZero
and his contribution to the field of deep reinforcement learning. Read more here:
Curious about AI and want to learn more? Introducing the DeepMind podcast!
@FryRsquared
will be taking you behind the scenes at DeepMind, exploring everything from AlphaGo to the future of AI. Read our blog, listen to the trailer and subscribe:
Exciting to see our most recent work on the dynamics of glassy systems published in
@Nature
today. Congrats to everyone at
@DeepMind
who contributed to understanding these complex systems!
DeepMind’s
@pushmeet
spoke to
@80000Hours
about the importance of building robust and safe AI systems.
Listen to the full podcast to find out why a career in machine learning makes a difference:
Great writeup on our engineering team's efforts that were critical to so many of our latest work and now ready to help accelerate ML research for the broader community.
We are excited to share the story behind TF-Replicator, a software library that helps developers deploy their TensorFlow code between CPUs, GPUs and TPUs with minimal effort.
Blog post:
Paper:
In our latest work, published in
@NatureMedicine
, we collaborated with
@Moorfields
&
@GoogleHealth
to train an AI system that could help predict the development of sight threatening age-related macular degeneration.
A decade since our founding & four years since the AlphaGo Challenge Match in Seoul, we caught up with
@bengeorges
to discuss why our next phase could be the most exciting yet.
Read more in today’s
@LesEchosWeekEnd
:
A big congratulations to the present, future and past DeepMinders appointed as CIFAR chairs at
#NeurIPS2019
today, in recognition of their outstanding contribution to the Canadian AI ecosystem
@CsabaSzepesvari
@cjmaddison
, Adam White & Aishwarya Agarwal
Congratulations Kathryn for this very well deserved recognition! Your contributions to
#AlphaFold
can't be understated - it is a great pleasure to work with you and the team.
Congratulations to Kathryn for being included in
@techreview
's
#35InnovatorsUnder35
list for her leadership & tremendous contribution to the development of
#AlphaFold
! We hope this work continues to accelerate research & pave the way toward new discoveries
I'm thrilled to announce I have joined
@DeepMind
to help their extraordinary team continue to lead in their quest to build AGI. More to come. I'm also so happy to reconnect with so many friends I worked with a decade ago. Let the adventure begin!
AI can deepen our knowledge of science & use that understanding to better the world. Yet the art of discovery lies in picking the right problems & questions to explore.
Science lead
@pushmeet
explains AI's role in asking powerful q's in
@sciam
here:
Effectively and efficiently deploying graph neural networks at scale is very challenging. Our team recently entered the Open Graph Benchmark Large-Scale Challenge & was one of the top-3 ‘winners’ on both topic classification & quantum chemistry tasks! (1/)
We're excited to see accelerating progress on rigorous specification-testing, robust training, and formal verification, from
@DeepMindAI
and from the broader ML community. If you’d like to work with us, we are hiring!
@uesatoj
@sgowal
Dj Dvijotham
Big congratulations to Karen Simonyan who has won the prestigious
#PAMI
Young Researcher Award for his contributions to the computer vision research community
#CVPR2019