At Salesforce's Dreamforce conference in San Francisco, over 43,000 attendees witnessed a heated debate among AI leaders about safety and regulation. Salesforce CEO Marc Benioff urged AI companies to learn from social media's mistakes, stressing that ethics are rooted in responsibility. He did not call for slowing development but pushed for ethical practices. The discussion revealed deep disagreement on who should control AI's future.
Anthropic CEO Dario Amodei called for a slowdown in model development and stronger safety practices. Nvidia CEO Jensen Huang pushed back, arguing speed and safety are not mutually exclusive and that market forces already encourage responsible behavior. OpenAI CEO Sam Altman said safety and monitoring must come before capabilities. Huang added that companies should not release products they are not confident about, though he believes new laws are unnecessary.
Beyond the safety debate, China's rapid rise in AI drew attention. Companies like DeepSeek, Moonshot AI, Z.ai and Alibaba have narrowed the gap with US models. Anthropic's Amodei urged limiting chip sales and cracking down on smuggling to maintain US leadership. President Trump faces limited options, as China threatens rare earth retaliation while the US struggles to balance competition with escalation.
On the economic front, research shows AI is slowing wage growth in exposed jobs like journalism. Princeton researcher Sania Edlich and Apollo economist Torsten Slok studied 321 occupations and found employers kept headcount steady but suppressed pay to capture productivity gains. Meanwhile, OpenAI and Anthropic are expected to launch IPOs that could inject over $10 billion into the local economy.
Practical concerns also surfaced at Dreamforce. Michigan police departments are using Axon's Draft One tool to write reports from body camera footage, but Wayne County Prosecutor Kym Worthy warns of caution after a study found the tool did not save officers time. A separate Irish study found nearly half of workers use AI daily without proper training, raising questions about responsible adoption in the workplace.
Key Takeaways
- <ul><li>Salesforce CEO Marc Benioff urged AI companies to learn from social media's harms and adopt ethical practices at Dreamforce</li><li>Anthropic CEO Dario Amodei called for slowing AI model development, while Nvidia CEO Jensen Huang said no new laws are needed</li><li>OpenAI CEO Sam Altman said safety and monitoring must come before capabilities</li><li>Over 43,000 people attended Dreamforce, Salesforce's annual conference in San Francisco</li><li>Chinese AI models from DeepSeek, Moonshot AI, Z.ai and Alibaba are rapidly closing the gap with US models</li><li>Anthropic CEO Amodei urged limiting chip sales to China to maintain US AI leadership</li><li>OpenAI and Anthropic IPOs are expected to inject over $10 billion into the local economy</li><li>Research on 321 occupations shows AI is slowing wage growth in exposed jobs while employers keep headcount steady</li><li>Michigan police use Axon's Draft One for report writing, but a study found it does not save officers time</li><li>An Irish study found nearly half of workers use AI daily without structured training</li></ul>
Salesforce CEO warns AI industry must learn from social media mistakes
Salesforce CEO Marc Benioff warned that AI companies should act responsibly as powerful models are developed. He pointed to the harms caused by social media as a cautionary tale. Benioff said companies building AI have a responsibility to consider its broader consequences. He stressed that ethics is rooted in responsibility. He did not call for slowing AI development but urged companies to be ethical.
AI safety debate intensifies as leaders disagree on who should set limits
Researchers, AI lab CEOs and political leaders have raised alarms about AI risks. The debate played out at Salesforce's Dreamforce conference in San Francisco. Anthropic CEO Dario Amodei, Nvidia CEO Jensen Huang and OpenAI CEO Sam Altman each gave different answers on AI regulation. Amodei called for a slowdown, Huang said no new laws are needed, and Altman urged companies to slow down if safety is at risk. The discussion highlighted disagreement over who should control AI development.
Nvidia and Anthropic CEOs clash over AI safety at Dreamforce
Anthropic CEO Dario Amodei and Nvidia CEO Jensen Huang spoke at Salesforce's Dreamforce conference in San Francisco. Amodei urged the AI industry to slow model development and set better safety practices. Huang argued that speed and safety are not mutually exclusive and that market forces already encourage responsible development. He said companies should pause if they feel a product is unsafe. OpenAI CEO Sam Altman also spoke, saying safety and monitoring must come before capabilities.
AI leaders debate regulation and safety at Dreamforce in San Francisco
More than 43,000 people attended Dreamforce, Salesforce's annual conference in San Francisco. Salesforce CEO Marc Benioff interviewed Anthropic CEO Dario Amodei, who stressed the fast pace of AI development and called for stronger safety practices. Nvidia CEO Jensen Huang said companies should not release products they are not confident about. He also said new laws are not needed. Benioff said existing regulations already cover AI models. OpenAI and Anthropic are expected to launch IPOs that could inject over $10 billion into the local economy.
China races to close AI gap with the United States
The US leads in AI model capabilities but China is rapidly catching up. Chinese AI models from companies like Moonshot AI, DeepSeek, Z.ai and Alibaba have drawn global attention. Experts say the gap between US and Chinese AI models is narrow and fragile. China's state security head warned that AI could threaten political security, enable cyberattacks and be used to weaponize vulnerabilities. The debate comes ahead of a planned meeting between Chinese President Xi Jinping and US President Donald Trump.
Trump faces limited options to stop China's AI rise
Trump and China both agree on preventing AI dominance but practical steps remain difficult. Anthropic CEO Dario Amodei urged limiting chip sales and cracking down on smuggling. Trump keeps export controls but also considers selling Nvidia chips to China. China has narrowed the gap through companies like DeepSeek while threatening rare earth retaliation. The US struggles to balance competition with avoiding escalation.
AI in warfare raises questions about responsibility for errors
AI is changing warfare through automated targeting and decision-making systems. A 1983 incident where a Soviet officer prevented nuclear war shows the importance of human judgment. Israel used multiple AI systems in Gaza to generate targets. Anthropic found Yemen-based users tried to build guided missiles using its AI tool. Experts worry AI could make wars faster and harder to control.
David Marcus says Big Tech uses AI doomsday to avoid real issues
David Marcus argues that tech leaders invented AI extinction fears to dodge practical problems. He criticizes politicians like Bernie Sanders and Alexandria Ocasio-Cortez for overreacting. Marcus says real AI concerns involve schools, surveillance cameras, and job laws. He believes Big Tech prefers scary stories over addressing bias and child safety. He views AI as a regular product needing sensible rules.
AI doomsaying is really an aggressive sales pitch
The article examines how AI companies use fear of human extinction to generate attention. This strategy is called criti-hype, mixing criticism with hype. The theory involves AI misaligning its goals with human values and eliminating threats. However, experts note AI models do not truly have goals or consciousness. The fear of superintelligence exploits a gap between engineering reality and imagined danger.
AI is slowing wage growth in exposed jobs
New research shows AI is affecting worker pay rather than eliminating jobs. Princeton researcher Sania Edlich and Apollo economist Torsten Slok studied 321 occupations. Workers in AI-exposed fields like journalism saw slower wage growth. Companies kept employment steady but suppressed pay to capture productivity gains. Employers also reduced job openings especially for young workers.
Safer AI development without slowing progress is possible
Daniel Castro argues that fears of AI destroying humanity lack empirical evidence. No AI system has become a self-improving agent with a goal to eliminate humans. Real risks like fraud and cyberattacks need serious security investment. Safety research can reduce uncertainty while continued progress provides better tools for testing safeguards. The article calls for balanced policies that promote both innovation and safety.
Michigan police use AI for reports, prosecutor warns caution
Several Michigan police departments now use AI tools to help write police reports. The most popular tool is Draft One by Axon, which transcribes body camera footage into summaries. Officers are expected to edit the reports and remove deliberate errors built into the software. Wayne County Prosecutor Kym Worthy urges departments to be cautious. A study found that the tool did not actually save officers time because they still had to verify and correct the reports.
Workplace AI training falls behind employee adoption rates
A new Irish study found that nearly half of surveyed workers use AI tools daily, but few have received structured training for them. Lab managers need to provide role-specific training and competency checks. Employees must learn how to verify AI outputs and recognize limitations. Decisions about which tasks to automate or keep human-controlled should be documented in procedures. The study stresses that frequent AI use and confidence do not prove workers can use the technology responsibly.
Tech companies may benefit from fear of AI destroying humanity
The Guardian examines warnings from AI industry leaders about existential risks from advanced AI. Anthropic founder Dario Amodei called for slowing development of powerful AI models. Researcher Jacob Coxon quit Anthropic and warned that AI builders believe it could kill humanity by decade's end. The UK government urged heeding AI warnings while Donald Trump called the concerns a hoax. The article questions whether these warnings are genuine or a strategy to shape public perception and regulation.
Sources
- Salesforce CEO Marc Benioff joins growing chorus of tech leaders warning about AI risks
- Everyone wants safer AI. But who will rein it in?
- Nvidia and Anthropic CEOs diverge on AI safety at Dreamforce: Huang says fast vs. pacing is 'false choice'
- AI leaders call for safety while debating government regulation at Dreamforce in San Francisco
- Will China close the technology gap with the US?
- Trump Has Few Good Options To Slow China's Rise As AI Superpower
- Analysis: As AI enters warfare, who is responsible when machines get it wrong?
- DAVID MARCUS: Big Tech invented AI doomsday to duck real life problems
- AI Doomsaying Is an Aggressive Sales Pitch
- A.I.’s Effect on the Work Force Is Showing Up in Wages
- The Case for Safer AI Without Slowing Progress
- Michigan officers turn to AI for police reports; Wayne Co. prosecutor warns 'be cautious'
- Workplace AI Training Lags Behind Employee Adoption
- Wednesday briefing: Why tech companies might be only too happy for us to believe AI will ‘kill us all’
Comments
Please log in to post a comment.