Download as PDF, PPTX
![11
Source: G. Gentry, IBM Research, Computing Arbitrary Functions of Encrypted Data – ACM
http://ece.gmu.edu/coursewebpages/ECE/ECE646/F10/project/F10_Project_Resources/Computing_Arbitrary_Functio
ns_of_Encrypted_Data.pdf
Usable Full Homomorphic Encryption
2010 2011 2012 2013 2014
Estimated amortized time
for computing a single bit
operation on encrypted data
Moore’s law
2009-10: Plausibility
[GH’11] A single bit
operation takes 30
minutes
IBM Patent 2013
2011-2012: “Real
Circuits”
[GHS’12] A 30,000-
gate in 36 hours
2013-today: Usability
HElib [HS’14]: IBM’s open-source FHE implementation
Basis for “generic” FHE computation
An “assembly language” for FHE
Implements Brakerski-Gentry-Vaikuntanathan (BGV) scheme
Security based on ring LWE (RLWE)
The same 30,000-gate circuit in 4-15 minutes
www.pieroleo.com](https://image.slidesharecdn.com/ibmresearch5in52018-180319114226/75/A-reading-of-the-IBM-Research-5-in-5-2018-Edition-11-2048.jpg)
![Assumptions for Quantum-Resilient Public-Key
Cryptography
N=pq
gx=y mod p
Factoring is hard
Finding short vectors in lattices is hard
Computing discrete logs is hard
(f,g) in
Z[x]/(xn+1)
Lattice-Based
Cryptography](https://image.slidesharecdn.com/ibmresearch5in52018-180319114226/75/A-reading-of-the-IBM-Research-5-in-5-2018-Edition-21-2048.jpg)
Uploaded byPietro Leo
A reading of the IBM Research 5-in-5 2018 Edition
Within five years, several emerging technologies will become mainstream and help address major challenges. Small autonomous AI microscopes deployed around the world will continually monitor plankton to anticipate ways to protect water supplies. Quantum computing will be used extensively to solve problems previously considered unsolvable, with IBM continuing to lead in this area. New solutions will counter a substantial rise in biased AI systems, with one method reducing bias in training data to minimize inequity in systems that learn from that data.
More Related Content
Similar to A reading of the IBM Research 5-in-5 2018 Edition
![[DSC Europe 23] Ales Gros - Quantum and Today s security with Quantum.pdf](https://cdn.slidesharecdn.com/ss_thumbnails/dsceurope23alesgros-quantumandtodayssecuritywithquantum-231128224122-311adb0d-thumbnail.jpg?width=640&height=640&fit=bounds)
A reading of the IBM Research 5-in-5 2018 Edition
- 1. 1 http://www.research.ibm.com/5-in-5/ Pietro Leo Executive Architect- IBM Italy CTO for Artificial Intelligence Chief Scientist for IBM Italy Research & Business IBM Academy of Technology Leadership Member of ISO/SC42 Artificial Intelligence Standardization Committee
- 2. IBM Research 5in 5
- 3. 3Quantum Computing Artificial Intelligence Biases AIRobotic Microscope Lattice Cryptography Crypto-anchors
- 4. 4 Nobody likes knockoffs. Crypto-anchors and blockchain willunite against counterfeiters. Crypto-anchors
- 5. 5 Crypto-anchors and blockchainwill unite against counterfeiters. Within five years, we’ll embed crypto- anchors such as ink dots and tiny computers smaller than a grain of salt in everyday products – everything from medical kits, to clothing, to car parts and food. We’ll put real-time information from them on a blockchain, giving us a foolproof way to know if a good is authentic and safe to consume. Crypto-anchors
- 6. 6 Crypto-anchors & Blockchain Source:https://www.youtube.com/watch?v=LUFWk55w5o0
- 7. 7 Preliminary scientific evidences IBMResearch crypto Anchors for Diagnostic devices: https://www.zurich.ibm.com/st/precision_diagnostics/cryptoanchors.html Crypto anchors extend blockchain’s Value Into the physical realm. Crypto Anchors address this challenge by embedding a security code in microfluidic diagnostic devices and “classical” lateral flow tests. This code can be used to identify and link products to highly secured digital transactions on the Cloud and using Blockchain. Crypto anchors come in many forms. They can even be embedded into an editable shade of a magnetic ink, which can be used to dye a malaria pill, for example IBM Research Crypto Anchor for diagnostic devices
- 8.
- 9. 9 We are preparingfor the day cyber- attackers will breach most of today’s cybersecurity methods and quantum computers will break all current forms of encryption. Our scientists have already submitted a post-quantum encryption method to the U.S. government for consideration as a potential standard. And they’ve developed new security methods based on lattice cryptography, which is essentially incredibly hard math. This technique makes it possible to perform calculations on a file without ever seeing sensitive data or exposing it to hackers. Lattice cryptography
- 10.
- 11. 11 Source: G. Gentry,IBM Research, Computing Arbitrary Functions of Encrypted Data – ACM http://ece.gmu.edu/coursewebpages/ECE/ECE646/F10/project/F10_Project_Resources/Computing_Arbitrary_Functio ns_of_Encrypted_Data.pdf Usable Full Homomorphic Encryption 2010 2011 2012 2013 2014 Estimated amortized time for computing a single bit operation on encrypted data Moore’s law 2009-10: Plausibility [GH’11] A single bit operation takes 30 minutes IBM Patent 2013 2011-2012: “Real Circuits” [GHS’12] A 30,000- gate in 36 hours 2013-today: Usability HElib [HS’14]: IBM’s open-source FHE implementation Basis for “generic” FHE computation An “assembly language” for FHE Implements Brakerski-Gentry-Vaikuntanathan (BGV) scheme Security based on ring LWE (RLWE) The same 30,000-gate circuit in 4-15 minutes www.pieroleo.com
- 12. 1. How toshare genomic data in a way that preserves the privacy of the data donors, without undermining the utility of the data or impeding its convenient dissemination? 2. How to perform a LARGE--SCALE, PRVIACY--PRESERVING analysis on genomic data, in an untrusted cloud environment or across multiple users? • Fast drop in the cost of genome- -sequencing 2000: $3 billion • Mar. 2014: $1,000 • Genotyping 1M variations: below $200 Unleashing the potenDal of the technology • Healthcare: e.g., disease risk detection, personalized • medicine • Biomedical research: e.g., • geno-phono • association • Legal and forensic • DTC: e.g., ancestry test, paternity test Full Homomorphic Encryption Examples – Genome Analysis
- 13. Source: The 2ndComparison on Critical Assessment of Data Privacy and Protect Secure Genome Analysis http://www.humangenomeprivacy.org/2015/slides/003_iDASH%20workshop%202015_setStage.pdf Computing the Hamming distance between two Human Genomes under a Full Homomorphic Encryption schema
- 15. https://arxiv.org/pdf/1512.04965.pdf “One of ourmain findings is that the number of logical qubits required to implement a Grover attack on AES is relatively low, namely between around 3000 and 7000 logical qubits.” “When realizing AES, only SubBytes Involves T-gates. Moreover, SubBytes is called a minimum of 296 times as in AES-128 and up to 420 times in AES-256. As shown above, for all three standardized key lengths, this results in quantum circuits of quite moderate complexity. So it seems prudent to move away from 128-bit keys when expecting the availability of at least a moderate size quantum computer.”
- 16. Source: https://www.iad.gov/iad/customcf/openAttachment.cfm?FilePath=/iad/library/ia-guidance/ia-solutions-for-classified/algorithm- guidance/assets/public/upload/CNSA-Suite-and-Quantum-Computing-FAQ.pdf Commercial NationalSecurityAlgorithm Suite 05 January 2016
- 17.
- 18. Source: https://pq-crystals.org/index.shtml Post-quantum cryptographysuite submitted by IBM and other partners to NIST call
- 19. Source: https://falcon-sign.info/ Post-quantum cryptographysuite submitted by IBM and other partners to NIST call
- 20. Source: IBM Researchercomments https://securityintelligence.com/preparing-next-era-computing-quantum-safe-cryptography/ N=pq gx=y mod p Factoring is hard Computing discrete logs is hard Finding short vectors in lattices is hard
- 21. Assumptions for Quantum-ResilientPublic-Key Cryptography N=pq gx=y mod p Factoring is hard Finding short vectors in lattices is hard Computing discrete logs is hard (f,g) in Z[x]/(xn+1) Lattice-Based Cryptography
- 22. 22 Our oceans are dirty.AI- powered robot microscopes may save them. AI Robotic Microscope
- 23. 23 In five years,small, autonomous AI microscopes, networked in the cloud and deployed around the world, will continually monitor the activities of plankton, which are natural, biological sensors of aquatic health. By observing certain indicators, we can anticipate ways to protect the global water supply. AI Robotic Microscope
- 24.
- 25. 25 Phytoplankton, Why arethey important? Surce: https://www.whoi.edu/main/topic/phytoplankton Phytoplankton are some of Earth's most critical organisms and so it is vital study and understand them. They generate about half the atmosphere's oxygen, as much per year as all land plants. Phytoplankton also form the base of virtually every ocean food web. In short, they make most other ocean life possible. Phytoplankton are mostly microscopic, single-celled photosynthetic organisms that live suspended in water.
- 26. 26 Simone BiancoIBM AlmadenResearch Center http://cellgeometry.ucsf.edu/sites/cellgeometry.ucsf.edu/files/STC_Project_5_CellStateInferenceEngine_final.pdf Cellular Sentinel Workflow
- 27. 27 From morphology vectorsto cell state Simone BiancoIBM Almaden Research Center http://cellgeometry.ucsf.edu/sites/cellgeometry.ucsf.edu/files/STC_Project_5_CellStateInferenceEngine_final.pdf
- 28. 28 Artificial Intelligence Biases AI biaswill explode. But only the unbiased AI will survive.
- 29. 29 Within five years,we’ll have new solutions to counter a substantial increase in the number of biased AI systems and algorithms. IBM researchers developed a method to reduce the bias that may be present in a training dataset, such that any AI algorithm that later learns from that dataset will perpetuate as little inequity as possible. Artificial Intelligence Biases
- 30. 30 Artificial Intelligence Biases Sorce:https://www.youtube.com/watch?v=q44XUZdCIMM
- 31.
- 32. Source: Gender Shades:IntersectionalAccuracy Disparities in Commercial Gender Classification, Buolamwin (MIT) and Gebru (Microsoft) http://proceedings.mlr.press/v81/buolamwini18a/buolamwini18a.pdf Face++ 34.5 (Dark Female) 0.7 (Dark Male) 6.0 (Light Female) 0.8 (Light Male) Microsoft 20.8 (Dark Female) 6.0 (Dark Male) 1.7 (Light Female) 0.0 (Light Male) IBM 34.7 (Dark Female 12.0 (Dark Male) 7.1 (Light Female) 0.3 (Light Male) Error Rates in gender classification Call to action 4 October 2017
- 33. Source: IBM MitigatingBias in AI Models - https://www.ibm.com/blogs/research/2018/02/mitigating- bias-ai-models/ Face++ (NO ANSWER) 34.5 (Dark Female) 0.7 (Dark Male) 6.0 (Light Female) 0.8 (Light Male) Microsoft (WILL ANDWER) 20.8 (Dark Female) 6.0 (Dark Male) 1.7 (Light Female) 0.0 (Light Male) IBM (ANSWERED - February 6) 3,46 (Dark Female) <- 34.7 1,99 Dark Male) <– 12.0 0,0 (Light Female) <– 7.1 0.25 (Light Male) <– 0.3 Public exposure of AI services is placing new kind of pressures to be solved quickly IBM (Production Env.- Update February 23)
- 34. Fairness, accountability andethics in AI § A novel probabilistic formulation of data pre-processing for reducing discrimination. Source: Calmon, Wei, Ramamurthy& Varshney, “Optimized Data Pre-Processing for Discrimination Prevention,” NIPS 2017
- 35. 35 Today, quantum computing isa researcher’s playground. In five years, it will be mainstream. Quantum Computing
- 36. 36 In five years,it will be mainstream. Quantum computing will be used extensively by new categories of professionals, developers, and students to solve problems once considered unsolvable. And given IBM’s longstanding leadership in quantum, they will continue to partner with us on future applications of this revolutionary technology. Quantum Computing
- 37.
- 38. 38 The future isquantum • A 20 qubits available on cloud to our clients at the end of this year (already made available 5-16 and 17 qubits) • A prototype with 50 qubits has been successful tested with a double of coherence time with respect previous systems. • An open source Quantum Information software (Qikit) • Quantum Computing: Breaking Through the 49 Qubit Simulation Barrier .Source: https://www.ibm.com/blogs/research/2017/11/the-future-is-quantum Source: https://www.ibm.com/blogs/research/2017/10/quantum-computing- barrier/
- 40. 40 Preliminary scientific evidences IBMQ Experience First QC on the cloud > 75,000 users All 7 continents > 2.8 million experiments 60+ external papers >150 colleges and universities
- 41. 41 Simulating molecules on quantumcomputers just got much easier with quantum hardware. https://www.ibm.com/blogs/research/2017/09/quantum-molecule/ Article published in Nature, Hardware-efficient Variational Quantum Eigensolver for Small Molecules and Quantum Magnets by IBM Research implement a new quantum algorithm capable of efficiently computing the lowest energy state of small molecules. Preliminary scientific evidences
- 42. 42 Example of externalpapers https://arxiv.org/pdf/1801.03782.pdf
- 43. IBM Research: Aglobal research capability Australia China Almaden Haifa Zurich Africa Ireland Brazil Watson Austin India Tokyo 3000 Researchers 6 Nobel Laureates 25 Years of Patent Leadership 10 National Medals of Technology 5 National Medals of Science 3 Kavli Prizes 6 Turing Awards 69 NAE Members 123 IEEE Fellows 28 ACM Fellows 99 IBM Fellows Cambridge Singapore
- 44. 44 http://www.research.ibm.com/5-in-5/ Pietro Leo Executive Architect- IBM Italy CTO for Artificial Intelligence Chief Scientist for IBM Italy Research & Business IBM Academy of Technology Leadership Member of ISO/SC42 Artificial Intelligence Standardization Committee @pieroleo www.pieroleo.com