Five for Friday: Issue #1
On Sutskever's rise from the ashes, Alibaba's Qwen2-VL model, 1X Technologies' NEO humanoid robot, the launch of Claude Enterprise, and the revamped Alexa
I'm excited to announce a new feature for New World Navigator called Five for Friday. Every Friday, I’ll be bringing you a punchy distillation and analysis of five of the most interesting and impactful AI news events from the past week.
From breakthrough innovations and key industry trends to the most talked-about news stories, Five for Friday will give you a quick, insightful overview of the stories shaping the future of AI, allowing you to stay informed and ahead of the curve.
#1 Rising from the Ashes
Ilya Sutskever’s Safe SuperIntelligence (SSI) Raises $1bn
Ilya Sutskever, who participated in the controversial and short-lived ousting of OpenAI’s CEO Sam Altman in November 2023, and who subsequently left the company in May 2024, made a big return to the AI scene this week.
The new start-up he cofounded, Safe Superintelligence (SSI) this week announced that it had raised over US$1bn (at a reported valuation of US$5bn) from a group of investors including Andreessen Horowitz, Sequoia Capital, DST Global, SV Angel and NFDG, an investment partnership co-run by SSI executive Daniel Gross.
SSI's aims to develop “safe” AI systems that surpass human capabilities and currently has 10 employees.
Perspectives:
Sutskever has made it clear that his sole aim is AGI, so we’ll unlikely see products from the company anytime soon, as Superintelligence or Artificial General Intelligence (AGI) is likely to be a medium- to long-term play.
While its heartening to witness the rise of a start-up whose core mission is safe and responsible AI, the same could have been said about OpenAI when it was initially founded. Only time will tell.
#2 Latest Kid on the Block
Alibaba releases new multimodal model Qwen2-VL
Researchers at Alibaba announced the release of Qwen2-VL, the latest iteration of models within the Qwen model family.
Qwen2-VL exhibited very strong visual capabilities across six dimensions including complex college-level problem-solving, mathematical abilities, document and table comprehension, multilingual text-image understanding, general scenario question-answering, video comprehension, and agent-based interactions.
The 72B model, available via API, excelled across a number of benchmarks, surpassing even GPT-4V and Claude 3.5-Sonnet on some measures. The smaller 2B and 7B models are available on an opensource basis via an Apache 2.0 license.
Perspectives:
The number of “GPT-4 class” models has grown over the past months, and Qwen2-VL is simply the latest addition to the roster. All eyes are now on OpenAI’s new model, codenamed Orion, to provide the next leap in AI capabilities.
On a separate note, the US’s efforts to constrain China’s AI research and industry do not appear to have slowed the latter’s technology giants from storming ahead.
#3 AI for Your Dishes and Laundry?
OpenAI-backed 1X Technologies unveils new humanoid robot
The Internet has been awash with memes of people lamenting that they would prefer AI to do their dishes and laundry rather than their art and writing. These folks need look no further!
1X Technologies, backed by OpenAI, unveiled its NEO Beta humanoid robot designed for household use. Unlike its competitors, NEO was designed from the group up to be a consumer robot and therefore has a soft exterior in case of collisions with humans.
Safety is a key consideration and the team will be deploying a number of NEOs into homes for testing and R&D later this year.
Perspectives:
While its clear that the future of our homes will involve robotics, its unclear yet if humanoid robots will be the solution. The benefit is that they present a familiar profile that could be helpful in certain situations (e.g., dealing with children), and of course human homes are built for human-shaped profiles. However, specialised non-humanoid robots may ultimately end up being more practical.
#4 Claude Goes Corporate
Claude launches its Enterprise proposition
Anthropic has launched its Claude Enterprise plan, offering organisations enhanced AI capabilities with enterprise-grade security.
Claude Enterprise features a 500K context window, allowing teams to work with extensive datasets, codebases, or documents, while maintaining security through SSO, role-based permissions, and audit logs. A native GitHub integration allows engineering teams to collaborate seamlessly on projects.
Early customers like GitLab and Midjourney have used Claude to streamline tasks such as coding, brainstorming, and research.
Perspectives:
As a big fan of Claude, I’m glad that they’ve finally joined the party with an Enterprise proposition. Given Claude’s reputation for responsibility and safety, its Enterprise proposition could appeal to organisations that are sensitive about this topic.
However, it is likely to be an uphill battle in the corporate market as Microsoft and to a lesser extent, OpenAI, have built up a significant lead.
#5 If You Can’t Beat Them, Co-Opt Them
Amazon turns to Claude for the revamped version of Alexa
Amazon is set to launch a revamped version of Alexa, powered by Anthropic’s Claude AI, in October 2024.
The ecommerce giant’s in-house AI model was rumoured to have far from satisfactory, leading to Alexa’s relaunch being delayed multiple times, and ultimately prompting the switch to Claude. “Remarkable Alexa” as the revamped version is known internally, will include features such as AI-generated news summaries, a child-focused chatbot, and conversational shopping tools.
Reports from earlier this year suggested that the company might price the revamped Alexa service at US$5-10 per month, which would diverge from the existing model where users pay only for hardware and not for the service.
Perspectives:
Amazon has struggled to keep pace with the likes of fellow tech giants Alphabet, Microsoft, and Meta in the AI race. While its AWS cloud computing division now dishes up a wide range of AI models, products and solutions, its consumer offering has been noticeably lackluster.
It will take more than a makeover of Alexa to bring the company back into the race, especially as consumers now have such a wide array of AI chatbots, including OpenAI’s Advanced Voice Mode for ChatGPT, Google’s Gemini voice chat, and Apple’s reimagined Siri to choose from.
Justin Tan is passionate about supporting organisations and teams to navigate disruptive change and towards sustainable and robust growth. He founded Evolutio Consulting in 2021 to help senior leaders to upskill and accelerate adoption of AI within their organisation through AI literacy and proficiency training, and also works with his clients to design and build bespoke AI solutions that drive growth and productivity for their businesses. If you're pondering how to harness these technologies in your business, or simply fancy a chat about the latest developments in AI, why not reach out?