Google DeepMind's CEO sees artificial general intelligence as a reality in ‘a few years’ — even as warnings of its potential dangers grow
Welcome to LinkedIn News Tech Stack, which brings you news, insights and trends involving the founders, investors and companies on the cutting edge of emerging technology.
As always, let me know if you have ideas or feedback on something that could be a fit by sending me an InMail. Follow me on LinkedIn for other tech updates.
Zooming In
A deep-dive into one big theme or news story every week.
Ever since LinkedIn parent Microsoft and OpenAI released a new paper on GPT-4, the AI world has been consumed by the concept of artificial general intelligence (AGI).
The research claimed that the algorithm powering GPT-4 was a nascent example of AGI, which OpenAI CEO Sam Altman describes as anything “generally smarter than humans.”
Now, the CEO of OpenAI competitor Google Google DeepMind has said that true artificial general intelligence could be less than a decade away. Speaking at The Wall Street Journal ’s Future of Everything summit on Tuesday, Demis Hassabis said that recent progress in AI had been “incredible” and is only going to accelerate.
“I don't see any reason why that progress is going to slow down,” he said. “I think we could be just a few years, maybe within a decade away.”
What is AGI?
While there’s no widely accepted definition of AGI, there is agreement that it’s when an AI system displays some kind of consciousness or sentience.
AGI — also sometimes known as strong AI — is when AI can potentially perform multiple complex intellectual tasks. For example, it can come up with a script, write the dialogue and create the video for a TV ad — going beyond performing a single function or task in the process.
In other words, it’s AI that’s as close to mimicking the human brain as possible. The brain, Hassabis said, is “the only example of general intelligence in the universe as far as we know.”
Executives like Altman and Hassabis have openly talked about their ambitions to develop the technology. Both OpenAI and DeepMind include the pursuit of AGI in their respective missions.
While OpenAI states its mission is “to ensure that artificial general intelligence benefits all of humanity,” DeepMind says its long-term aim is “to solve intelligence, developing more general and capable problem-solving systems, known as artificial general intelligence (AGI).”
The AI community is divided on its promise
AGI, however, has become a thorny issue among the AI community. As the companies above lead the charge to develop it, some are dismissing it as science-fiction.
“GPT-4 and Auto-GPT have many limitations as AI systems,” Aakash Gupta , a product growth executive who most recently worked at Affirm, wrote on LinkedIn. “They do not have a deep or broad understanding of the meaning or context of what they generate. They do not have a robust or flexible ability to reason, learn, or adapt beyond their training data. They do not have a clear or reliable way to explain their outputs or correct their mistakes. These are some of the essential features of artificial general intelligence (AGI).”
Others appear almost fearful. Tesla CEO Elon Musk has called it “a danger to the public,” while author and theoretical physicist Stephen Hawking once declared that an AGI “could spell the end of the human race.”
Now, a growing faction of AI researchers and practitioners is issuing warnings about the humanitarian consequences of investment in AGI that goes unchecked, without guardrails in place.
Just this week, for instance, Geoffrey Hinton, a pioneer of artificial intelligence, quit Google so that he could speak freely about the technology's dangers. He said he feared that computers could become smarter than people far sooner than he and other experts had predicted. Altman himself has warned of AGI's "massive risks."
Most agree on developing AGI with caution, but differ on the approach
While in the short term, advancements in AGI could lead to the disruption of jobs, in the long term, it has far-reaching implications concerning privacy, social equity and energy resources. Those could be an obstacle for developing countries in particular.
Even The White House seems concerned, and has summoned the chief executives of Google , Microsoft , OpenAI and Anthropic to discuss key AI issues this week, including the "expectation that companies like yours must make sure their products are safe before making them available to the public,” Reuters reported.
For their part, the companies at the forefront of developing the technology appear to be on board, at least for now.
Hassabis said that part of his new job was to integrate AI developments into Google products, and he seemed to be in favor of adopting a tempered approach.
“I would advocate developing these types of AGI technologies in a cautious manner using the scientific method, where you try and do very careful, controlled experiments to understand what the underlying system does,” he said.
Greg Brockman , the co-founder and president at OpenAI, has a different take. He recently made the case that the best way to develop AGI ethically was to put the technology out in the public, as opposed to behind the closed doors of research labs.
“That default plan of being, well, you build in secret, you get this super powerful thing, and then you figure out the safety of it and then you push ‘go’ and you hope you got it right — I don't know how to execute that plan,” he said during a recent TED talk. “For me, that was always terrifying.”
Brockman continued: “I think that this alternative approach is the only other path that I see, which is that you do let reality hit you in the face. And I think you do give people time to give input. Before these machines are perfect, before they are super powerful, you actually have the ability to see them in action.”
This Week in AI
Here’s where we bring you up-to-speed with the latest advancements from the world of AI.
- As mentioned above, Geoffrey Hinton, widely regarded as the “godfather” of machine learning, has resigned from Google to be able to speak freely about the potential dangers of AI. Hinton joined Google in 2013 and is responsible for some of the biggest breakthroughs in artificial intelligence.
- Here are more details on the White House’s meeting with tech leaders: The White House has announced a $140 million investment for seven new artificial intelligence research hubs ahead of a meeting with Big Tech leaders Thursday. Alphabet Inc. ’s Sundar Pichai, OpenAI’s Sam Altman, Anthropic’s Dario Amodei and Microsoft’s Satya Nadella are set to meet with Vice President Kamala Harris and other officials. On the agenda: A "frank discussion" of AI's risks and how to ensure the technology is developed in a safe, trustworthy way. The White House has proposed an AI Bill of Rights that discusses transparency, data privacy and other concerns, while the Commerce Department has outlined a risk management framework.
- Salesforce today announced a new generative AI tool called SlackGPT, aimed at increasing productivity in its workplace communication tool by leveraging its own and other third-party LLMs. The company also announced that Einstein GPT, its generative AI assistant, will be coming on Slack as a dedicated app soon. “Our goal is to help our customers to connect better with their customers – and AI represents a step-function in that direction,” Patrick Stokes , EVP & GM of platform at Salesforce, told me in an interview. “We have an incredible opportunity to drive incremental value.”
- LinkedIn parent Microsoft announced another significant expansion of its AI-based search tools today, rolling out new features that allow visual and multimodal searches and persistent chat tools. These features expand the capabilities of its search engine Bing and its web browser Edge.
- LinkedIn co-founder Reid Hoffman and DeepMind co-founder Mustafa Suleyman have rolled out a new chatbot called Pi, which is being billed as “a kind and supportive companion that’s on your side.” The chatbot is developed by the company Inflection AI, which says that Pi is a supportive “neutral listener” that can talk with users about personal problems and offer advice.
- IBM said it will wait and see which jobs can be replaced by artificial intelligence before hiring for more back-office roles. Roughly 7,800 jobs in departments such as human resources could be done with AI and automation, CEO Arvind Krishna says. A few days ago, Dropbox too announced that it would use AI to handle the work formerly done by roughly 500 employees.
- Samsung Electronics has banned the use of generative AI tools like ChatGPT a month after internal, sensitive data from the company was accidentally leaked.
- AI was a hot topic during Universal Music Group and Spotify’s earnings calls this week. Spotify CEO Daniel Ek said it was important to balance creative innovation while protecting artists on the platform, adding that AI would lead to more content from people who don’t play instruments, resulting in more revenue. Universal Music Group CEO Lucian Grainge echoed Ek’s message about the burden of enforcing copyright rules, though he contends AI could help with immersive sound, as well as mixing and master recordings.
- A coalition of technology and educational organizations like Khan Academy , the World Economic Forum and Educational Testing Service are teaming up with Code.org to support TeachAI, an effort to bring AI to the classroom and help educators explain how the technology works to their students.
- PETA has given the Bible’s Book of Genesis a modern makeover using ChatGPT for its latest stunt, peppering the book with animal rights messages and vegan teachings. The company hopes that its “Story of Creation” will appeal to Gen Z, 73% of whom identify as animal rights activists.
- And finally, U.S. respondents of LinkedIn’s ongoing Workforce Confidence Index survey were recently asked which digital skills they would like to gain or improve this year. While the most popular answer was data analysis (with 28% selecting that option), AI and machine learning skills came next, with a quarter of workers wanting to upskill in that area. See more here.
TechTok
Catch up on the tech headlines you may have missed this week and what our members are saying about them on LinkedIn.
- Bluesky Social is having a moment. The invite-only social media app, co-founded by former Twitter CEO Jack Dorsey , has begun to lure an influential and especially vocal crowd of tech media figures and influencers. The app doubled its new user base in a single day late last week.
- The Federal Trade Commission on Tuesday announced plans to ban Meta from monetizing the data of under-18s, saying the company had “repeatedly violated its privacy promises” and “put young users at risk.” Regulators said Meta failed to meet the terms of its $5 billion privacy settlement from 2020, while Meta called the FTC proposal a “political stunt” and vowed to fight the ban.
- Tired of racking your brain after another password change? Google knows that, and it's rolling out an alternative. The tech giant has introduced passkeys, a feature it calls "a major step" toward eliminating passwords for good. Users will be able to access their accounts — much like they might access their phone — by scanning their finger or face, or using a PIN.
- Apple and Google may be rivals, but they agree on at least one thing: Location-tracking devices such as Apple's AirTags can pose privacy risks. The tech firms have proposed a new industry standard that would alert Android and iOS users “in the event of suspected unwanted tracking.”
- TikTok users may soon notice a new feature on the app, geared toward stemming the onslaught of AI-driven deepfakes. The social media company is working on a specialized label that will allow creators to mark if their content was made with the use of AI, The Information reports, citing unnamed sources. Snap is similarly "considering how to identify manipulated videos," according to The Information.
- Twitter has restored free access to its backend interface for public entities after its paid version caused apps to break and organizations to quit the platform. The site introduced a three-tiered pricing model for its API in March that charged organizations from $100 a month for a basic model to tens of thousands for enterprise-level tiers.
Movers and Shakers
Here’s keeping tabs on key executives on the move and other big pivots in the tech industry. Please send me personnel moves within emerging tech.
- Atlassian has hired former Instagram and Facebook exec Hemal Shah as a VP of product.
- Snap Inc. has hired former Meta exec Patrick Harris as SVP of partnerships and Dave Sommer , who most recently served as chief commercial officer at Fetch, as its head of verticals, as it looks to bolster its ad business.
- TikTok ’s head of U.S. trust and safety, Eric H. , is leaving the company, according to The Verge.
- Ann Lai , who was in charge of a data-driven diligence program at VC firm Bullpen Capital , announced in a LinkedIn post that she had been fired as a general partner at the firm, saying she believed the move was both discriminatory and retaliatory.
- VC firm Catalio Capital Management has hired the executive chairman of Haystack Oncology, Isaac Ro , as a partner.
- Venture firm Greycroft has hired Kevin Gasque , previously at The Carlyle Group Group, as CFO and COO.
- Stormbreaker Ventures has hired Glenn Lurie , ex-CEO of AT&T mobility and consumer ops, as a partner. The firm also hired serial entrepreneur Bart Lorang as a venture partner.
ICYMI
Here are other top stories of the week from beyond LinkedIn in the broader world of tech.
- Once-booming audio app Clubhouse is cutting more than half of its staff.
- Meta has flagged more than 1,000 domains since March that are distributing malware-laced ChatGPT-themed tools, according to a new report.
- Mozilla has acquired Fakespot by Mozilla , a small startup whose website and browser plug-in help users identify bogus product reviews on e-commerce sites.
- Amazon has a new TikTok competitor. Insider lists everything you need to know.
Thanks for reading. Please share Tech Stack and forward it around if you like it! And if you have any news tips, find me on InMail.
Disability
1yWhat is that trying to say I bought just a apple gift card
--
1yThe future start and begin in here an go far beyond expectations and the rest we do not know anything
Absolutely! The integration of AGI with LinkedIn's tech stack is an exciting development, as it has the potential to improve user experience and provide more personalized and relevant content.
Empowering HR & Change Leaders to Drive Successful Change | Boost Engagement & Innovation for Sustainable Growth | 16 Yrs. Supporting Organisational Change & Professional Transitions | Executive Coach, Trainer
1yFi Hills Missak Vehouni Collette Stone MSc MBA Chartered FCIPDAbigail M. Mike Harsanyi Rajib Chowdhury