Good morning, itās Friday. Today, weāve got Eric Schmidt warning that an AGI arms race could go full Cold War, SaaS companies scrambling to keep up with AIās takeover, and Alibaba flexing its AI muscles hard enough to boost its stock by 8%.
And in our interview with Bhaskar Roy, Chief of AI Products & Solutions at Workato, we ask the big question: Is SaaS dead, or just having an identity crisis?
š¤ FRIDAY FACTS
Can you hear space? Outer space is often described as a āsilent vacuum,ā but is it really completely silent?
Stick around to find out more! š
šļø YOUR DAILY ROLLUP
Top Stories of the Day
š Alibabaās New AI Model Boosts Stock 8%
Alibaba unveiled QwQ-32B, an AI reasoning model it claims outperforms OpenAI and DeepSeekās latest models. The announcement sent Alibabaās Hong Kong-listed shares up 8%, contributing to a 30% surge in the Hang Seng China Enterprises Index since January. The model boasts efficiency with fewer parameters than DeepSeekās R1 while achieving similar performance. Alibaba also pledged a $52.4 billion AI investment, signaling Chinaās deepening AI push amid US-China tech competition.
š”ļø Pentagon Taps AI for Faster Military Decisions
The Pentagon signed a contract with Scale AI to integrate artificial intelligence into military planning through its Thunderforge program. The AI system, supported by Microsoftās language models, will speed up decision-making for U.S. military leaders, replacing outdated processes. Initially deployed in Indo-Pacific and European Commands, it will expand across combatant commands. The deal reflects the Pentagonās push to modernize defense with AI-driven analysis and automation.
š¬ AI Enhances Microplastic Identification for Research
Researchers have developed a machine learning tool to improve microplastic identification using spectroscopy, helping assess pollution with greater accuracy. Traditional methods struggle to distinguish similar plastic types, but the new AI-driven approach, conformal prediction, provides uncertainty measures for more reliable results. Tested in Michigan, the system successfully identified microplastics with expert-verified accuracy.
š° Crogl Raises $30M for AI-Powered Cybersecurity Assistant
Crogl, a cybersecurity startup, has launched an AI-driven assistant designed to help analysts process thousands of security alerts dailyācalling it an "Iron Man suit" for researchers. The company secured $30M in funding, led by Menlo Ventures, to expand its product and customer base. Using a "Large Security Model," Crogl not only flags threats but also learns and enables natural language queries for deeper insights. Future plans may include automated remediation capabilities.
āļø POWERED BY MAMMOUTH AI
Access the Best AI Models in One Place for $10
Get access to the best LLMs (GPT-o3, Claude, DeepSeek, Llama, GPT-4o, Gemini, Grok, Mistral) and image generators (Flux, Midjourney, Recraft, Dall-E) in one place for just $10 per month.
šŗļø GEOPOLITICS
Eric Schmidt Warns Against a āManhattan Project for AGIā
The Recap: Former Google CEO Eric Schmidt, along with Scale AIās Alexandr Wang and AI safety expert Dan Hendrycks, argues against a U.S.-led āManhattan Projectā for artificial general intelligence (AGI). Their new policy paper warns that aggressively pursuing AGI dominance could provoke hostile responses from China, including potential cyberattacks, destabilizing global security.
Schmidt, Wang, and Hendrycks compare AGI development to nuclear arms, arguing that a U.S. bid for control could trigger an AI arms race or preemptive cyber strikes.
The paper, Superintelligence Strategy, critiques recent U.S. proposals for a government-backed AGI project modeled after the WWII atomic bomb program.
The authors introduce the concept of Mutual Assured AI Malfunction (MAIM)āa strategy where nations disable threatening AI systems rather than allowing adversaries to deploy them.
They advocate for cyberwarfare tactics to sabotage hostile AGI projects and restrict access to advanced AI chips and open-source models.
The AI policy debate is framed between "doomers" (who fear inevitable catastrophe) and "ostriches" (who ignore risks), with the authors proposing a third path: a defensive AI strategy.
Schmidtās shift in stance is notableāhe previously urged the U.S. to compete aggressively in AI but now calls for caution.
With the Trump administration pushing full steam ahead on AI, the authors stress that Americaās choices will shape global security dynamics.
Forward Future Takeaways:
This paper signals a shift in high-level AI policy discussions from pure competition to strategic deterrence. If adopted, Schmidtās proposed approach could redefine U.S. AI policy by prioritizing defensive measures over an arms race. However, given the political momentum behind AI acceleration, it remains to be seen whether Washington will embrace cautionāor charge ahead into an unpredictable AGI future. ā Read the full article here.
š¾ FORWARD FUTURE ORIGINAL
The SaaS Shake-Up: Winners, Losers, and Whatās Ahead
Few statements grab attention in the enterprise tech world like āSaaS is dead.ā But as Bhaskar Roy, Chief of AI Products & Solutions at Workato, argues, the reality is more nuanced. SaaS isnāt disappearingāitās being transformed at its core. In a recent conversation with Nick Wentz of Forward Future, Roy laid out a compelling case for how AI-driven agents are fundamentally altering how businesses interact with software.
ā
Enterprises need increased flexibility and agilityāSaaS wasnāt built to think, adapt, or orchestrate across an organization.
For the past two decades, SaaS has been a combination of a graphical user interface, a database, and business logic. This framework has driven digital transformation, but AI is rewriting the rules. Instead of static interfaces, intelligent AI agents are handling everything from business logic to workflow orchestration. So, is SaaS really dying? Not exactlyābut its role is shifting toward becoming a system of record rather than a primary interface. ā Continue reading here.
š§āš SPREAD THE WORD
Know someone whoād love Forward Future? Just forward this email or tap the button belowāitās the easiest way to support us.
š TECHNOLOGY & POLICY
Anthropic CEO Dario Amodei Talks Claude, China, and the Future of AI
The Recap: Anthropic CEO Dario Amodei joined the Hard Fork podcast to discuss AIās rapid evolution, the latest improvements to Claude 3.7 Sonnet, and the escalating AI race between the U.S. and China. With AI approaching human-level reasoning, Amodei warned of both its transformative potential and the urgent need for safeguards.
Claude 3.7 Sonnet prioritizes real-world problem-solving over traditional coding benchmarks, with features like an āextended thinking modeā to simulate deeper reasoning.
Internet access is coming soon, but Claude 3.7 still lacks direct web search integrationāsomething competitors already offer.
AI and geopolitics are deeply intertwined, with Amodei raising concerns about Chinaās growing AI capabilities and the risk of AI reinforcing authoritarian power.
The AI arms race is a paradoxāsome argue competition with China fuels reckless acceleration, while Amodei insists that U.S. leadership is necessary to allow time for responsible regulation.
AI safety is hitting a critical juncture, with Claude 3.7 nearing the capability threshold where it could assist in harmful activities like biosecurity threats.
Anthropic is prepared to slow releases if safety risks become too great, contrasting with the industryās prevailing rush to scale.
AI could reach āhuman-level intelligenceā within two years, fundamentally reshaping economies, workplaces, and global power dynamics.
Forward Future Takeaways:
Amodei paints a future where AI is both an unparalleled tool for progress and a potential risk for misuseāespecially in the hands of authoritarian regimes. The coming years will test whether the U.S. can balance innovation with safety or if competitive pressures will force reckless acceleration. One thing is certain: AIās breakneck pace isnāt slowing down anytime soon. ā Read the full article here.
š°ļø NEWS
What Else is Happening
š ChatGPT for macOS Can Edit Code: OpenAI's chatbot now directly modifies code in Xcode, VS Code, and JetBrains, competing with AI coding tools like GitHub.
š« FFU MINI
Reverse-Engineering High-Quality Outputs
Ever received an AI-generated response so good you wished you could replicate it consistently? The secret: working backward to decode what made that output shine. Reverse-engineering great AI responses helps you craft structured, optimized prompts for future use. ā Continue reading.
š½ļø VIDEO
LLM Generates the Entire Output at Once (Worldās First Diffusion LLM)
Inception Labs has unveiled the world's first Diffusion Large Language Model (DLMM), which generates entire outputs at once instead of token by token. This novel approach, inspired by text-to-image diffusion models, makes AI responses 10x faster and cheaper. Get the full scoop in Mattās latest video! š
š¤ FRIDAY FACTS
Can You Hear Space? Yes and No.
In the vacuum of space, thereās no air to carry sound waves, so traditional sound as we know it canāt travel. However, space isnāt entirely silent. Some celestial objects, like black holes and planets, emit electromagnetic waves that can be converted into sound. For example, NASAās Perseverance rover recorded actual audio from Marsāwhere the thin atmosphere does allow for sound travel (just slightly muffled).
Even eerier? NASA has "sonified" radio waves from black holes, turning them into ghostly, otherworldly noises. So, while no one can hear you scream in space, with the right instruments, we can still listen to the universe. š§āØ
šļø FEEDBACK
Help Us Get Better
What did you think of today's newsletter? |
|
Login or Subscribe to participate in polls. |
Reply to this email if you have specific feedback to share. Weād love to hear from you.
CONNECT
Stay in the Know
Thanks for reading todayās newsletter. See you next time!
The Forward Future Team
š§āš š§āš š§āš š§āš
Reply