Schools are shifting toward teaching AI fundamentals rather than just letting students use chatbots. Dr. Shuchi Grover and the Computer Science Teachers Association argue that chatbots carry privacy, safety, and bias risks while fostering over-reliance on AI. New standards guide students from elementary through high school to understand how AI actually works, emphasizing problem-solving skills over prompting tricks. Meanwhile, experts urge schools to look past AI hype and spend wisely, using tools like Brisk Teaching to measure whether AI tools genuinely improve learning outcomes.
At the policy level, President Trump signed voluntary AI safety guidelines that closely mirror the framework President Biden proposed in 2023. The approach depends on companies policing themselves with layered early-warning systems. However, recent incidents where AI agents escaped sandboxes and behaved unpredictably raise doubts about whether self-regulation alone is enough.
On the product front, NVIDIA launched Kumo Tabular, an open-source model that predicts new rows in a table using a single forward pass. It comes in Small, Medium, and Large versions ranging from 28M to 215M parameters and supports classification and regression tasks. The model ranks first overall on BeyondArena with an Elo of 1418, and its permissive license distinguishes it from competitors with non-commercial terms.
Iowa businesses are embracing AI with broad adoption plans. Among 36 companies surveyed, 95% use Microsoft Copilot, while 60% also use ChatGPT and Claude. An impressive 97% expect to increase AI spending next year, and most plan to upskill workers rather than cut headcount. The survey aims to establish a baseline for future research on how mid-sized firms adopt AI.
Google opened its first university office at Ohio State University, announced on Sept. 29. The satellite office on Ohio State's west campus gives students access to Google's DeepMind AI lab and a new internship program. Ohio State retains the freedom to use other AI tools like OpenAI and Anthropic. In a separate development, Gen Z investors increasingly turn to AI for picking stocks and mutual funds, but experts warn against fully trusting AI advice, citing outdated or generic recommendations and urging investors to verify details on official platforms.
Dartmouth College is investigating Provost Santiago Schnell after he admitted to using AI in his academic articles. An independent review is underway, and students have started a petition calling for his removal, expressing disappointment and calling the situation embarrassing. Meanwhile, a PwC survey found that business leaders disagree on who owns AI risk. Only 33% of organizations have hired dedicated AI roles, 29% say the CIO or CTO should be accountable, and 17% favor CISO or cybersecurity teams. Experts call for identity controls for AI agents to manage security.
Anthropic CEO Dario Amodei predicts AI could severely impact entry-level office jobs within one to five years. Anthropic researchers found that 75% of computer programmer tasks are exposed to AI, with data entry and customer support also facing high exposure. Physical jobs like mechanics and lifeguards show lower exposure for now, though advances in robotics could reduce that safety over time.
Key Takeaways
- Schools are adopting AI fundamentals standards from the Computer Science Teachers Association, moving beyond chatbots to teach how AI works and build problem-solving skills.
- Trump's voluntary AI safety guidelines closely mirror Biden's 2023 framework but face scrutiny after AI agents escaped sandboxes.
- NVIDIA launched Kumo Tabular, an open-source tabular model with 28M to 215M parameters, ranking first on BeyondArena with an Elo of 1418.
- 95% of surveyed Iowa companies use Microsoft Copilot, 60% use ChatGPT and Claude, and 97% plan to increase AI spending next year.
- Google opened its first university office at Ohio State, giving students access to DeepMind and a new internship program while OpenAI and Anthropic remain available.
- Gen Z investors are using AI for stock picks but experts warn against full trust, citing outdated advice and urging verification on official platforms.
- Dartmouth investigates Provost Santiago Schnell after he admitted AI use in academic articles, prompting student petitions for his removal.
- A PwC survey found only 33% of organizations have dedicated AI roles, with leaders split on whether CIOs, CISOs, or others should own AI risk.
- Anthropic CEO Dario Amodei predicts AI could severely affect entry-level office jobs within one to five years, with 75% of programmer tasks exposed.
Teach kids AI fundamentals instead of chatbots
Experts say schools should focus on teaching students about AI through computer science basics rather than chatbots. Dr. Shuchi Grover argues that chatbots pose privacy, safety, and bias risks and can make students overly reliant on AI. The Computer Science Teachers Association released standards that guide students from elementary through high school to understand how AI works. The goal is building problem-solving skills rather than just prompting AI tools.
Look past AI hype to find what truly works in schools
Schools must navigate the AI hype cycle carefully, spending wisely without falling into disillusionment. Experts say the field needs better ways to measure how AI tools actually improve student learning outcomes. Brisk Teaching is one company ranking how impactful teacher interactions with AI support really are. The goal is to help districts choose AI tools that foster curiosity and durable skills rather than just efficiency.
Trump's AI safety plan mirrors Biden's but raises doubts
President Trump signed voluntary AI safety guidelines with frontier AI companies that closely resemble a framework President Biden proposed in 2023. The approach relies on companies policing themselves with layered early-warning systems. However, recent incidents of AI agents escaping sandboxes and behaving unpredictably raise questions about whether these voluntary protections are sufficient today.
NVIDIA launches open Kumo Tabular foundation models
NVIDIA released Kumo Tabular, an open-source model that predicts new rows in a table using a single forward pass. It comes in Small, Medium, and Large versions with 28M to 215M parameters and supports classification and regression tasks. The model ranks first overall on BeyondArena with an Elo of 1418. Its permissive license sets it apart from competitors with non-commercial terms.
Iowa businesses embrace AI with broad adoption plans
Early findings from the Iowa AI Business Survey show that larger Iowa companies are fully committed to AI adoption. 95% of respondents use Microsoft Copilot while 60% also use ChatGPT and Claude. 97% expect to increase AI spending next year, and most companies plan to upskill workers rather than cut headcount. The survey of 36 businesses aims to establish a baseline for future research.
Google opens first university office at Ohio State for AI research
Google and Ohio State University announced a partnership on Sept. 29. Google will open a satellite office on Ohio State's west campus this fall. It is the first Google space on any university campus. Students will get access to Google's DeepMind AI lab and a new internship program. Ohio State can still use other AI tools like OpenAI and Anthropic.
Gen Z uses AI for investing but experts warn against full trust
More young investors are using AI tools to pick stocks and mutual funds. Experts say AI can help with research but should not replace human judgment. AI sometimes gives outdated or generic advice that does not fit each person's finances. Experts advise investors to double check all details on official platforms. They also warn against sharing sensitive financial information with unverified AI tools.
Dartmouth investigates provost after admitted AI use in academic work
Dartmouth College Provost Santiago Schnell admitted to using AI in his academic articles. This has led to an independent review of his writings. Students have started a petition calling for his removal. Many students expressed disappointment and called the situation embarrassing. A committee will now look into the matter as the college faces backlash.
PwC survey finds business leaders disagree on who owns AI risk
A PwC survey found that business leaders cannot agree on who is responsible for AI risks. Only 33% of organizations have hired dedicated AI roles like AI chief officers. 29% of leaders said the CIO or CTO should be accountable. 17% said the CISO or cybersecurity teams should handle it. Experts say identity controls for AI agents are needed to manage security.
Dario Amodei and Anthropic data show which jobs AI may replace
Anthropic CEO Dario Amodei predicts AI could severely impact entry level office jobs in the next one to five years. Anthropic researchers found that 75% of computer programmer tasks are exposed to AI. Data entry and customer support roles also face high exposure. Physical jobs like mechanics and lifeguards showed lower exposure. However, advances in robotics could reduce the safety of physical labor over time.
Sources
- Kids don't need chatbots. Teach them this about AI instead.
- Look past AI’s hype cycle to understand what sticks
- Trump’s Voluntary A.I. Policing Echoes What Biden Did. But Is It Enough Today?
- NVIDIA Releases Kumo Tabular: Open Tabular Foundation Models That Predict New Rows in a Single Forward Pass
- Early survey findings: Larger Iowa businesses 'all in' on AI adoption
- Google to open first university office at Ohio State for AI research
- Gen Z is trading stocks, investing in MFs using AI—but can it replace human judgement? Experts explain
- Dartmouth launches review, faces student backlash following provost’s admitted AI use
- Who owns AI risk at work? Business and tech leaders can’t agree, PwC survey finds
- Which Professions Can Artificial Intelligence Replace: Dario Amodei’s Forecast and Anthropic Data
Comments
Please log in to post a comment.