OpenAI is actively reshaping its strategy by pursuing new partnerships with Hiro and TBPN. These moves appear designed to strengthen ChatGPT's presence in enterprise settings, with industry observers viewing the transactions as targeted acqui-hires to secure specific talent. While the deals are modest relative to OpenAI's massive scale, they signal a willingness to experiment with different acquisition paths to maintain competitiveness.
In the workforce sector, Verizon CEO Dan Schulman is pushing for greater transparency regarding AI-driven job cuts. He established a $20 million fund to assist employees in transitioning careers, acknowledging that admitting to layoffs is difficult for many leaders. Schulman warns that unemployment could rise to 20 to 30 percent within the next few years as AI and humanoid robots increasingly replace manual labor roles.
The entertainment industry is also adapting to these shifts. The thriller movie Bitcoin: Killing Satoshi is utilizing generative AI to slash production costs, avoiding a potential $300 million price tag for filming 200 distinct locations. Despite the technology, the project retained 200 crew members, including 55 AI artists dedicated to post-production work.
Security experts are raising alarms about the new risks AI introduces. Netskope's Tony Burnside highlights that east-west traffic generated by AI tools creates a fresh security frontier, where context poisoning could inject malicious data into communications. Meanwhile, the NSA continues to deploy Anthropic's Mythos AI model despite Pentagon warnings, limiting access to roughly 40 organizations to mitigate offensive cyber risks.
On the innovation front, a secondary market for internal data from failed startups is emerging. Companies like SimpleClosure are licensing Slack threads and emails from defunct firms to train reinforcement learning models. This data offers unique insights into organizational decision-making, though privacy advocates caution that anonymization efforts may fail to fully protect employee identities.
Key Takeaways
- OpenAI entered deals with Hiro and TBPN, likely as acqui-hires to bolster enterprise capabilities for ChatGPT.
- Verizon CEO Dan Schulman launched a $20 million fund to help workers transition due to anticipated AI job losses.
- Verizon predicts unemployment could reach 20-30% in 2-5 years as AI and robots displace manual laborers.
- The movie Bitcoin: Killing Satoshi used AI to avoid a $300 million production cost for filming 200 locations.
- 55 AI artists were employed on the Bitcoin film to handle post-production tasks alongside 200 human crew members.
- OpenAI CTO Srinivas Narayanan is leaving the company to return to India after three years.
- The NSA is using Anthropic's Mythos AI model despite Pentagon concerns about supply chain reliability.
- Anthropic has restricted Mythos access to roughly 40 organizations to prevent misuse in cyber operations.
- A secondary market for internal startup data is emerging, with SimpleClosure licensing Slack threads for AI training.
- Privacy advocates warn that employee data in failed startups may remain identifiable even after anonymization.
OpenAI explores new deals with Hiro and TBPN
OpenAI recently entered into deals with the personal finance startup Hiro and the media startup TBPN. These moves suggest the company is trying to refocus on making its ChatGPT models competitive in enterprise contexts. Experts describe these transactions as potential acqui-hires aimed at bringing talent on board. The deals are considered small compared to OpenAI's overall scale but show a willingness to try different things.
Verizon CEO urges transparency on AI job losses
Verizon CEO Dan Schulman told CEOs to be open about artificial intelligence-related job cuts in an interview with The Wall Street Journal. He launched a $20 million career-transition and retraining fund to help workers during this difficult time. Schulman noted that while change is necessary, it can be hard for companies to admit job losses. He recommended that employees use AI to write obituaries or poems to understand how their lives might change.
Movie Bitcoin uses AI for low-cost production
The thriller movie Bitcoin: Killing Satoshi is being filmed using generative AI to reduce production costs. The film stars Gal Gadot and Pete Davidson and follows an investigation into the creator of bitcoin. Director Doug Liman chose AI because practical filming of 200 distinct locations would have cost over $300 million. The movie still employed 200 crew members, including 55 AI artists to handle post-production tasks.
Verizon predicts high unemployment from AI and robots
Verizon CEO Dan Schulman warned that unemployment could climb to 20 to 30 percent in the next two to five years due to AI and humanoid robots. He urged other CEOs to give employees fair warnings about these inevitable changes in the workplace. Schulman created a fund to help workers transition as they enter the age of AI. He believes robots will eventually usurp many manual laborers in the American workforce.
Netskope highlights AI security visibility challenges
Tony Burnside from Netskope explained that visibility is key to securing AI systems in organizations. He noted that east-west traffic created by AI tools working across systems is a new security front. Burnside warned about risks like context poisoning where malicious data is injected into AI communications. He emphasized that security controls must be omni-directional to protect sensitive information without degrading performance.
OpenAI CTO Srinivas Narayanan leaves for India
Srinivas Narayanan, the CTO of B2B Applications at OpenAI, announced he will leave the company after three years. He plans to return to India to spend time with his aging parents before deciding his next steps. Narayanan led engineering teams behind ChatGPT and enterprise tools during his tenure. He thanked CEO Sam Altman and president Greg Brockman for the incredible journey.
Failed startups sell internal data for AI training
A secondary market has emerged for internal data from failed startups including Slack threads and emails. Companies like SimpleClosure now license this data to AI developers for building reinforcement learning gyms. This data provides structured information on how decisions unfold inside organizations which is hard to replicate with public data. Privacy advocates warn that employee data remains identifiable even after anonymization efforts.
NSA uses Anthropic AI despite Pentagon warnings
The US National Security Agency is using Anthropic's Mythos AI model even though the Pentagon labels the company a supply chain risk. The NSA continues to deploy the model while the Department of Defense questions Anthropic's reliability in court. Anthropic has limited access to Mythos to roughly 40 organizations to prevent misuse in offensive cyber operations. The NSA is believed to use the model to identify vulnerabilities in its own systems.
Sources
- OpenAI’s existential questions
- Verizon's Dan Schulman Tells CEOs to Be Open About AI Job Cuts
- Bitcoin Thriller Billing Itself As First Quality AI Generated Movie
- Verizon CEO: AI Is Coming for Your Job—'and Everyone Knows It'
- Netskope's Tony Burnside - visibility is key to AI security
- Indian-Origin OpenAI CTO Srinivas Narayanan Quits, Plans To Spend Time With "Ageing Parents In India"
- Slack chats and internal data from failed startups are finding a second life in AI training
- US security agency adopts Anthropic AI amid Pentagon’s ‘supply chain risk’ warning
Comments
Please log in to post a comment.